Perioperative Medicine

E   Meta-Analysis

Effectiveness of Perioperative Opioid Educational

Initiatives: A Systematic Review and Meta-Analysis
Andres Zorrilla-Vaca, MD,* Gabriel E. Mena, MD,† Pedro T. Ramirez, MD,‡ Bradley H. Lee, MD,§‖
Alexandra Sideris, PhD,§ and Christopher L. Wu, MD,§‖
See Article, p 938
BACKGROUND: Opioids are the most commonly prescribed analgesics in the United States.
Current guidelines have proposed education initiatives to reduce the risk of chronic opioid
consumption, yet there is lack of efficacy data on such interventions. Our study evaluates the
impact of perioperative opioid education on postoperative opioid consumption patterns includ-
ing opioid cessation, number of pills consumed, and opioid prescription refills.
METHODS: The MEDLINE/PubMed, Embase, Cochrane Library, Scopus, and Google Scholar data-
bases were systematically searched for randomized controlled trials (RCTs) assessing the impact
of perioperative educational interventions (using either paper- or video-based instruments regard-
ing pain management and drug-induced side effects) on postoperative opioid patterns compared to
standard preoperative care among patients undergoing elective surgery. Our end points were opioid
consumption (number of pills used), appropriate disposal of unused opioids, opioid cessation
(defined as no use of opioids), and opioid refills within 15 days, 6 weeks, and 3 months.
RESULTS: In total, 11 RCTs fulfilled the inclusion criteria, totaling 1604 patients (804 received
opioid education, while 800 received standard care). Six trials followed patients for 15 days after
surgery, and 5 trials followed patients up to 3 months. After 15 days, the opioid education group
consumed a lower number of opioid pills than those in the control group (weighted mean difference
[WMD], −3.39 pills; 95% confidence interval [CI], −6.40 to −0.37; P =.03; I2 = 69%) with no sig-
nificant difference in overall opioid cessation (odds ratio [OR], 0.25; 95% CI, 0.04–1.56; P = .14;
I2 = 83%). Likewise, perioperative opioid education did not have significant effects on opioid ces-
sation at 6 weeks (OR, 0.69; 95% CI, 0.45–1.05; P = .10; I2 = 0%) and 3 months (OR, 0.59; 95%
CI,0.17–2.01; P = .10; I2 = 0%) after surgery, neither reduced the need for opioid refills at 15 days
(OR, 0.57; 95% CI, 0.28–1.15; P = .12; I2 = 20%) and 6 weeks (OR, 1.08; 95% CI, 0.59–1.98; P =
.80; I2 = 37%). There was no statistically significant difference in the rate of appropriate disposal of
unused opioids between both groups (OR, 1.99; 95% CI, 0.66–6.00; P = .22; I2 = 71%). Subgroup
analysis by type of educational intervention showed a statistical reduction of opioid consumption
at 15 days when implementing multimedia/audiovisual strategies (4 trials: WMD, −4.05 pills; 95%
CI, −6.59 to −1.50; P = .002; I2 = 45%), but there was no apparent decrease when using only
paper-based strategies (2 trials: WMD, −2.31 pills; 95% CI, −12.21 to 7.59; P = .65; I2 = 80%).
CONCLUSIONS: Perioperative educational interventions reduced the number of opioid pills con-
sumed at 15 days but did not demonstrate a significant effect on opioid cessation or opioid
refills at 15 days, 6 weeks, and 3 months. Further randomized trials should focus on evidence-
based educational interventions with strict homogeneity of material to draw a more definitive
recommendation.  (Anesth Analg 2022;134:940–51)

KEY POINTS
• Question: What is the impact of perioperative opioid education on opioid consumption pat-
terns after hospital discharge (ie, number of pills, cessation, and prescription refills)?
• Findings: There was a significant reduction in the number of opioid pills consumed within 15
days after surgery among those patients who received opioid education, but there was no
impact on opioid cessation or refills within 15 days, 6 weeks, and 3 months.
• Meaning: Perioperative educational interventions may reduce opioid consumption after hos-
pital discharge, but it does not impact cessation or prescription refills.

From the *Department of Anesthesiology, Perioperative and Pain Medicine, Conflicts of Interest: See Disclosures at the end of the article.
Brigham and Women’s Hospital, Boston, Massachusetts; Departments of
Supplemental digital content is available for this article. Direct URL cita-
†Anesthesiology and Perioperative Medicine and ‡Gynecologic Oncology
tions appear in the printed text and are provided in the HTML and PDF ver-
& Reproductive Medicine, The University of Texas MD Anderson Cancer
sions of this article on the journal’s website (www.anesthesia-analgesia.org)
Center, Houston, Texas; §Department of Anesthesiology, Hospital for Special
Surgery, New York, New York; and ‖Department of Anesthesiology, Weil Reprints will not be available from the authors.
Cornell Medicine, New York, New York.
Address correspondence to Andres Zorrilla-Vaca, MD, Department of
Accepted for publication April 26, 2021. Anesthesiology and Perioperative Medicine, The University of Texas MD
Funding: None. Anderson Cancer Center, Houston, TX. Address e-mail to andres.zorrilla@
correounivalle.edu.co.
Copyright © 2021 International Anesthesia Research Society
DOI: 10.1213/ANE.0000000000005634

940 www.anesthesia-analgesia.org May 2022 • Volume 134 • Number 5

Copyright © 2021 International Anesthesia Research Society. Unauthorized reproduction of this article is prohibited.
 E  Meta-Analysis

GLOSSARY
CI = confidence interval; ERAS = Enhanced Recovery After Surgery; FDA = Food and Drug
Administration; GRADE = Grading of Recommendations, Assessment, Development, and Evaluation;
OR = odds ratio; ORADEs = opioid-related adverse drug events; PRISMA = Preferred Reporting
Items for Systematic Reviews and Meta-Analyses; PROs = patient-reported outcomes; RCT = ran-
domized controlled trials; SD = standard deviation; WMD = weighted mean difference

P
atients undergoing surgery are frequently lww.com/AA/D578. Boolean operators were used
exposed to opioids during hospitalization and to adapt the search strategy based on the database.
continue to take these medications after dis- In addition, citations were reviewed to ensure inclu-
charge.1 In the United States, approximately two- sion of relevant studies not captured in our initial lit-
thirds of ambulatory patients are prescribed opioids at erature search. Two authors reviewed and extracted
discharge,2 and more than half of these prescriptions the data independently (AZV and BHL). Full-text
remain unused.3 This phenomenon has contributed to articles that met inclusion criteria were reviewed
a national opioid crisis with mental health and eco- for detailed comprehension and further assessment
nomic burdens for the health care system and adverse of the quality and risk of bias. We excluded trials
consequences on patient outcomes including risks assessing education interventions in nonsurgical
of addiction, misuse, dependency, and diversion.4 patients. All disagreements between reviewers in the
As a result, institutions have undertaken strategies selection and evaluation processes were resolved by
to combat this crisis by minimizing opioid prescrip- discussion with a third reviewer (CLW). All demo-
tions,5 using opioid-sparing analgesic modalities, and graphic data, including year of publication, sample
implementing educational interventions.6,7 Patient size, type of educational strategy, type of surgery,
education is a process of influencing patient behavior and specified outcomes, were abstracted to a pre-
that will ultimately produce changes in knowledge, defined proforma.
attitudes, and skills.
Although perioperative opioid education and Eligibility Criteria
counseling have demonstrated benefits in terms of This meta-analysis was limited to randomized con-
patient satisfaction, safe behaviors, and psychologi- trolled trials (RCTs) among adult patients (age >18
cal outcomes,8 most systematic reviews have been years) who underwent surgery and were allocated
limited to assessing qualitative evidence.9 Some to either a perioperative opioid education interven-
clinical trials have demonstrated the effectiveness of tion or conventional education. For the purpose of
perioperative opioid education in reducing opioid this meta-analysis, we considered the perioperative
use after hospital discharge10,11; however, other tri- education defined by each trial, which mainly con-
als have yielded varying results and warrant further sisted of informative sessions about the risks of opi-
evaluation of the literature.12,13 The goal of our study oid dependency and the use of alternatives for pain
was to evaluate the impact of education on opioid control. However, the type of perioperative educa-
consumption, opioid cessation and opioid refills, and tion varied among the studies. We did not restrict
appropriate disposal of unused opioids after hospital our selection criteria to studies in specific regions
discharge. of the world or reporting outcomes at any specific
time interval. We included peer-reviewed original
METHODS articles.
This meta-analysis was conducted following the
Preferred Reporting Items for Systematic Reviews Intervention
and Meta-Analyses (PRISMA) statement.12 It is reg- Any perioperative opioid education strategy (ie, ver-
istered in the International Prospective Systematic bal, written, video, or any combination) was consid-
Reviews Registry database. ered as the intervention as long as the description
included information about opioid-related adverse
Literature Search Strategy drug events (ORADEs), pain modulation, and risks
MEDLINE/PubMed, Embase, Cochrane Library, of opioid overuse. We included studies in which the
Scopus, and Google Scholar databases were searched intervention was conducted during hospitalization
from inception to September 2020 for randomized either before or after surgery.
studies, assessing the impact of perioperative opi-
oid education on opioid consumption after hospital Outcomes
discharge. Our search strategy is explained in detail The primary outcome in this meta-analysis was the
in Supplemental Digital Content, File, http://links. number of opioid pills consumed within 15 days, 6

May 2022 • Volume 134 • Number 5 www.anesthesia-analgesia.org 941

Copyright © 2021 International Anesthesia Research Society. Unauthorized reproduction of this article is prohibited.
 Perioperative Opioid Education and Opioid Consumption

weeks, and 3 months after surgery. Secondary outcomes significant heterogeneity, respectively). Publication
included rate of opioid cessation, number of patients bias was calculated using the Begg and Egger test.15
who needed opioid refills, and appropriate disposal of Funnel plots were constructed to represent any ten-
unused opioids (defined as any of the proper Food and dency for publishing in favor of positive effect.
Drug Administration [FDA]–approved disposal meth- Significant publication bias was considered when
ods including dropoff at a drug takeback site such as there was asymmetry in the funnel plot, and a statisti-
hospitals, pharmacies, and police stations).14 cally significant bias coefficient was noted according
to the Egger test.15 P values <0.05 were considered
Methodological Quality and Quality of Evidence as statistically significant in all statistical analyses.
Methodological quality assessment was performed All analyses were performed using a random-effect
using the Cochrane Risk of Bias tool for random- model (DerSimonian & Laird method) given the high
ized studies. Each study was assessed based on the 7 clinical heterogeneity.17 All statistics were performed
domains of potential bias (random sequence genera- using Stata version 14.0 (Stata, College Station, TX).
tion, allocation concealment, blinding of intervention,
blinding of outcome assessment, incomplete outcome RESULTS
data, selective reporting, and other bias). The quality Study Characteristics
of the evidence provided in this meta-analysis was An initial search yielded 412 articles. After exclud-
assessed using the Grading of Recommendations, ing duplications, we screened a total of 289 titles
Assessment, Development, and Evaluation (GRADE) and abstracts. A total of 271 articles were excluded
criteria, as previously reported.15 The GRADE recom- due to the lack of education intervention or the
mendations were ranked based on the quality and assessment of education interventions in nonsurgi-
level certainty (ie, risk of bias, inconsistency, indirect- cal patients. Two additional studies were excluded
ness, and imprecision) of the available evidence. due to lack of randomization. Ultimately, 16 full-text
articles were reviewed, of which 1 additional study
Statistical Analysis was included after cross-checking references and 5
The data were initially tabulated and represented studies were excluded as they did not assess opioid
using descriptive statistics. An exploratory qualita- consumption patterns after hospital discharge. In
tive analysis was conducted to describe the charac- total, 11 RCTs fulfilled the inclusion criteria, which
teristics of the studies included in this meta-analysis. comprised 1604 patients (804 received opioid educa-
The risk of opioid cessation was extracted as a dichot- tion, while 800 received standard care).10–14,17–22 Seven
omous variable (present or absent) and compared trials were conducted in orthopedic surgery, 2 in
using odds ratios (ORs) with their respective 95% breast surgery, 1 in urogynecology, and 1 in obstet-
confidence intervals (CIs). We used forest plots to rical surgery (cesarean section). Five trials followed
illustrate the estimations and overall effect sizes patients until 15 days after surgery, 6 trials reported
with pooled OR represented as a solid diamond at information for 6 weeks after surgery, and 3 followed
the bottom of the forest plot. Outcomes presented patients until 3 months (Table 1). All trials started the
as continuous variables (eg, number of opioid pills) educational intervention preoperatively, and 5 trials
were compared using weighted mean difference also complemented the intervention with discharge
(WMD). In the cases of publication of median val- education. Figure 1 shows the PRISMA flowchart of
ues with their ranges, we converted these measures trial selection.
into mean and standard deviations (SDs) using the
method by Wan et al.16 In the cases of publication of Outcomes
95% CI of mean values, we calculated the SD using Opioid Pill Consumption. Six trials reported
the formula n*
( upper − lower limits of 95% CI ) . information regarding opioid pills consumption at
3.92 15 days. Three trials showed a statistically significant
Predetermined subgroup analyses were performed reduction of opioid pills consumption in the group
based on the material used for education (video, that received perioperative education. Meta-analysis
written, or both). Interaction statistics between the revealed a statistically significant reduction in opioid
subgroups were calculated to validate any potential pills consumed within the first 15 days that favored
difference in the results. Sensitivity analysis was per- perioperative opioid education interventions (WMD,
formed based on overall study quality (high or moder- −3.39 pills; 95% CI, −6.40 to −0.37; P = .03; I2 = 69%;
ate versus low) as determined by quality of evidence Figure 2). There was not enough information to assess
assessment. We also used the leave-one-out strategy to the effects on the reduction of opioid pills at 6 weeks
assess the consistency of our results. Heterogeneity (I2) and 3 months. In subgroup analysis by type of surgery,
was assessed with the correspondent χ2 test (I2 < 50% opioid consumption at 15 days was still reduced when
and I2 > 50% were considered insignificant and using perioperative opioid education among patients

942   
www.anesthesia-analgesia.org ANESTHESIA & ANALGESIA
Copyright © 2021 International Anesthesia Research Society. Unauthorized reproduction of this article is prohibited.
 Table 1. Description of the Clinical Trials Included in This Meta-Analysis
Sample Method for observing
E  Meta-Analysis

Study size Intervention Surgery Outcomes Follow-up outcome

Darnall et al,12 I: 36 Video-based instrument of 90 min with information and skills to regulate Breast Opioid cessation 15 d, 6 wk, REDCap survey (Inter-
2019 C: 32 cognition, emotion, and physiologic hyperarousal related to pain, including 3 mo net) or phone calls
relaxation, thought reframing, and behaviors that modulate attention and
counteract helplessness about pain
Egan et al,16 I: 39 Paper-based instrument divided into 3 sections. It contained information Breast Opioid pills consumed and 15 d Internet survey

May 2022 • Volume 134 • Number 5

2020 C: 46 about pain expectations, goals of tolerable pain, risks associated with opioid refills
opioid use, adjunct medications that may be used perioperatively, non-
medication pain control methods, and statements to normalize the pain
experience for the patient
Lam et al,17 I: 84 Video-based education instrument with handouts and in-person instruction Cesarean Opioid cessation, opioid pills 15 d Phone calls
2020 C: 91 about optimal analgesic use consumed, and refills
Stepan et al,11 I: 93 Webinar and paper-based education discussing information about periopera- Hand surgery Opioid pills consumed and 15 d Phone calls
2020 C: 97 tive pain management, expected pain after surgery, side effects of opioids, opioid refills
tiered system of pain management, and modalities of pain relief
Syed et al,10 I: 68 Video-based instrument of 2 min containing a handout detailing the risks Rotator cuff repair Opioid cessation and opioid 6 wk, 3 mo Follow-up visits or
2018 C: 66 of narcotic overuse and abuse. In addition, patients received education pills consumed phone calls
detailing recommended postoperative opioid usage, side effects, depen-
dence, and addiction.
Bloom et al,13 I: 65 Video-based educational instrument on the proper use and dangers of Rotator cuff repair Opioid refills 6 wk, 3 mo Phone calls and elec-
2020 C: 65 opioids tronic records
Nahhas et al,14 I: 183 Paper-based instrument with educational pamphlets plus text messages on Arthroplasty Opioid cessation and opioid 6 wk Follow-up visits and
2020 C: 139 proper disposal of unused opioids refills electronic records
Sabesan,18 I: 25 Paper-based educational materials on alternative pain management proto- Shoulder arthroplasty Opioid cessation 15 d Phone calls and
et al 2020 C: 25 cols, as well as discharge instructions detailing the opioid crisis, postop- follow-up visits
erative alternative nonopioid medications, and pain management plan.
Patients were also counseled verbally by a member of the research team
on pain expectations after surgery, nonopioid alternatives to control pain,
and over-the-counter analgesics for pain relief.
Ilyas et al,20 I: 107 Video-based educational instrument about the nature of opioids, statistics on Shoulder, elbow, and Opioid pills and refills 15 d Follow-up visits and
2021 C: 130 the “Opioid Epidemic” in America, how to safely consume opioids, nonopi- wrist surgeries phone calls
oid pain management strategies, and how to avoid opioid dependence
Singh et al,19 I: 39 Written instructions outlining postoperative pain expectations and Foot and ankle surgery Opioid refills 6 wk Phone calls
2018 C: 39 recommendations for opioid medication usage and disposal
Buono et al,21 I: 65 Paper-based educational pamphlets containing information regarding side Reconstructive pelvic Opioid pills, refills, and 15 d, 6 wk Follow-up visits and
2021 C: 70 effects and risks of opioid consumption, safe storage patterns of opioid surgery disposal phone calls
tablets, and recommendations regarding postoperative opioid consumption
Abbreviations: C, control; I, intervention; REDCap, Real Time Electronic Digitally Controlled Analyzer Processor.

www.anesthesia-analgesia.org
943
Copyright © 2021 International Anesthesia Research Society. Unauthorized reproduction of this article is prohibited.
 Perioperative Opioid Education and Opioid Consumption

who underwent orthopedic surgery (3 trials; WMD, consumption at 15 days when implementing multi-
−5.11 pills; 95% CI, −7.25 to −2.97; P = .001; I2 = 0%), but media/audiovisual strategies (4 trials: WMD, −4.05
there was no significant reduction in nonorthopedic pills; 95% CI, −6.59 to −1.50; P = .002; I2 = 45%), but
surgery (3 trials: WMD, −1.12 pills; 95% CI, −5.29 to there was no apparent decrease when using only
−3.06; P = .60; I2 = 65%) (ie, cesarean, urogynecology, paper-based strategies (2 trials: WMD, −2.31 pills;
and breast surgeries). Test for interaction was 95% CI, −12.21 to 7.59; P = .65; I2 = 80%). Test for inter-
statistically significant (P = .002), meaning that, based action was also statistically significant (P = .012).
on the current evidence, the effect of perioperative
opioid education is effectively greater in orthopedic Opioid Cessation. Six trials reported information about
population than in nonorthopedic populations. opioid cessation. Three trials included information at
Subgroup analysis by type of educational inter- 15 days, 3 at 6 weeks, and 3 at 3 months after surgery.
vention showed a statistical reduction of opioid Perioperative opioid education did not have significant

Figure 1. PRISMA flowchart of the selection of studies. PRISMA indicates Preferred Reporting Items for Systematic Reviews and Meta-Analyses.

944   
www.anesthesia-analgesia.org ANESTHESIA & ANALGESIA
Copyright © 2021 International Anesthesia Research Society. Unauthorized reproduction of this article is prohibited.
 E  Meta-Analysis

Figure 2. Forest plot showing the positive impact of perioperative opioid educational interventions on the reduction in the number of opioid
pills consumed within 15 d after surgery. CI indicates confidence interval; WMD, weighted mean difference.

impact on opioid cessation at 15 days (OR, 0.25; 95% Figure 1, http://links.lww.com/AA/D578. Overall,
CI, 0.04–1.56; P = .14; I2=83%), 6 weeks (OR, 0.69; 95% 6 studies were classified as low risk, and 3 stud-
CI, 0.46–1.06; P = .10; I2 = 0%), or 3 months (OR, 0.59; ies as moderate risk of bias. The sensitivity analy-
95% CI, 0.17–2.01; P = .10; I2 = 55%) after surgery. sis demonstrated the consistency of our results
Forest plots for this outcome are shown in Figure 3. It (Supplemental Digital Content, Figure 3, http://
was not possible to conduct subgroup analysis for this links.lww.com/AA/D578).
outcome due to the low number of trials.
GRADE Ranking Recommendations
Opioid Refills. The number of patients who required Based on the current evidence and level of certainty
opioid refills was reported in 4 trials at 15 days and 6 (Table 2), the effectiveness of perioperative education
weeks. There was no difference in opioid refills between on the reduction of opioid pills consumption after
both groups at 15 days (OR, 0.57; 95% CI, 0.28–1.15; hospital discharge is moderate. There was low rec-
P = .12; I2 = 20%; Figure 4) and 6 weeks (OR, 1.08; 95% ommendation for the outcomes: opioid cessation and
CI, 0.59–1.98; P = .80; I2 = 37%; Figure  4). There was need for refills, and very low for appropriate disposal
not enough information for opioid refills at 3 months. It of unused opioid pills.
was not possible to conduct subgroup analysis for this
outcome due to the low number of trials. Publication Bias
We were able to draw an Egger plot for opioid pills
Appropriate Disposal of Unused Opioid Pills. Two out of consumption (Supplemental Digital Content, Figure 4,
3 trials showed a statistically significant improvement http://links.lww.com/AA/D578), which did not
in the disposal of unused opioids. After pooling all 3 show significant publication bias (−2.06, P = .37). It
trials, there was no statistically significant difference was not possible to construct funnel plots for the rest
in the rate of appropriate disposal of unused opioids outcomes due to the low number of studies.
between both groups (OR, 1.99; 95% CI, 0.66–6.00;
P = .22; I2 = 71%; Supplemental Digital Content, Figure DISCUSSION
2, http://links.lww.com/AA/D578). In this meta-analysis, our group summarized the
most updated RCTs evaluating the effectiveness of
Methodological Quality Assessment perioperative opioid education interventions on opi-
The assessment of study quality and the risk for oid consumption over time, cessation, and refills after
bias are shown in Supplemental Digital Content, hospital discharge. Our results showed a statistically

May 2022 • Volume 134 • Number 5 www.anesthesia-analgesia.org 945

Copyright © 2021 International Anesthesia Research Society. Unauthorized reproduction of this article is prohibited.
 Perioperative Opioid Education and Opioid Consumption

Figure 3. Nonsignificant effects of opioid educational interventions on postoperative opioid cessation. Impact of opioid education on postop-
erative opioid cessation at 15 d (A), 6 wk (B), and 3 mo (C). CI indicates confidence interval; OR, odds ratio.

significant impact of opioid education interventions requirement of opioid refills when implementing opi-
on the reduction of postoperative opioid pills con- oid education interventions.
sumption at 15 days postsurgery. However, there Perioperative opioid education has become
was no reduction in the risk of opioid cessation and increasingly more relevant over the last few decades.

946   
www.anesthesia-analgesia.org ANESTHESIA & ANALGESIA
Copyright © 2021 International Anesthesia Research Society. Unauthorized reproduction of this article is prohibited.
 E  Meta-Analysis

Figure 4. Forest plot for the number of patients who needed opioid refills. Impact of perioperative opioid education on opioid refills at 15 d
(A) and at 6 wk (B). CI indicates confidence interval; OR, odds ratio.

The role of perioperative education interventions postoperative pain control. Other observational stud-
has been reported widely in the literature.19 There is ies have shown improved awareness about the correct
evidence that patient education initiatives improve disposal of opioid medications.29,30
postoperative patient satisfaction23 and increase the The effectiveness of opioid education interven-
level of knowledge on pain management among tions in decreasing postoperative opioid consumption
surgical patients.24 Furthermore, simple interven- could be explained by increased patient engagement
tions such as preoperative pamphlets or nursing and a better understanding regarding the role of
education may reduce acute pain after surgery and nonopioid analgesic modalities, pain expectations,
improve functional outcomes.19,25 The long-term ben- and the potential risk of opioid-related side effects.
efits remain unknown given the controversial results Based on Table 1, these 3 were the most common top-
among the studies assessing the effectiveness of such ics included in the education interventions and seem
interventions.26 to contribute to the reduction in opioid consump-
Our findings are consistent with large observa- tion. As noted in this meta-analysis, there is a wide
tional studies. Although this meta-analysis demon- variety of educational strategies. Interestingly, opioid
strated a statistically significant reduction of 3–4 pills pills consumption was reduced when implement-
at 15 days when implementing perioperative opi- ing audiovisual/multimedia methods but not when
oid education, this effect is not clinically significant. using only paper-based strategies, which suggest
Individual cohort studies suggest that the presence that patient engagement may improve when using
of opioid education initiatives may be associated more interactive educational initiatives. Other recent
with a decrease in postdischarge opioid prescribing studies have also suggested preoperative behavioral
while improving disposal after surgery.24,27 Hite et al28 modalities for education.31 My Surgical Success is a
implemented an opioid education initiative over a digital behavioral intervention that has been applied
1-year period and demonstrated the feasibility of this among patients undergoing breast surgery and has
specific type of intervention without compromising shown positive impact on the risk of opioid cessation

May 2022 • Volume 134 • Number 5 www.anesthesia-analgesia.org 947

Copyright © 2021 International Anesthesia Research Society. Unauthorized reproduction of this article is prohibited.
 948   
Table 2. GRADE Recommendations and Summary of the Evidence
Sample size Effect
Studies Risk of bias Inconsistency Indirectness Imprecision Opioid education Control Relative (95% CI) Certainty

www.anesthesia-analgesia.org
Opioid pill consumption (follow-up: median 15 d; assessed with: number of pills)
6 Not serious Serious Not serious Not serious 518 570 WMD, −3.39 pills lower (−6.40 lower to −0.37 lower) ⊕ ⊕ ⊕◯
MODERATE
Opioid cessation (follow-up: median 15 d; assessed with: patients who ceased opioids)
Perioperative Opioid Education and Opioid Consumption

3 Not serious Serious Not serious Serious 23/145 (15.9%) 50/148 (33.8%) OR, 0.25 (95% CI, 0.04–1.56) ⊕ ⊕◯◯
LOW
Opioid cessation (follow-up: median 6 wk; assessed with: patients who ceased opioids)
3 Not serious Serious Not serious Serious 48/219 (21.9%) 47/171 (27.5%) OR, 0.69 (95% CI, 0.46–1.06) ⊕ ⊕◯◯
LOW
Opioid cessation (follow-up: median 3 mo; assessed with: patients who ceased opioids)
3 Not serious Serious Not serious Serious 9/101 (8.9%) 8/97 (8.2%) OR, 0.59 (95% CI, 0.17–2.01) ⊕ ⊕◯◯
LOW
Opioid refills (follow-up: 15 d; assessed with: need for opioid refills)
4 Not serious Serious Not serious Serious 18/342 (5.3%) 35/374 (9.4%) OR, 0.57 (95% CI, 0.28–1.15) ⊕ ⊕◯◯
LOW
Opioid refills (follow-up: 6 wk; assessed with: need for opioid refills)
4 Not serious Serious Not serious Serious 116/352 (32.9%) 90/301 (29.9%) OR, 1.08 (95% CI, 0.59–1.98) ⊕ ⊕◯◯
LOW
Appropriate disposal of unused opioids (assessed with: patients who disposed unused opioids appropriately)
4 Not serious Serious Not serious Very serious 80/241 (33.2%) 26/206 (12.6%) OR, 2.61 (95% CI, 0.84–8.04) ⊕ ◯◯◯
VERY LOW
Abbreviations: CI, confidence interval; GRADE, Grading of Recommendations, Assessment, Development, and Evaluation; OR, odds ratio; WMD, weighted mean difference.

ANESTHESIA & ANALGESIA

Copyright © 2021 International Anesthesia Research Society. Unauthorized reproduction of this article is prohibited.
 E  Meta-Analysis

and patient satisfaction.12 Other common types of each outcome may increase the likelihood of type I
institutional strategies involve educating clinicians error in our findings, though there was no evidence
regarding prescribing behavior (ie, procedure-spe- of publication bias in our primary outcome. Second,
cific opioid-prescribing guidelines, and departmental it was not possible to analyze pain scores after hos-
policy.32–34 Those 2 interventions should complement pital discharge due to the different presentation of
patient education to obtain significant benefits in opi- the results between the trials. Third, currently, there
oid consumption and greater level of knowledge on is not a validated opioid educational intervention;
multimodal-pain control.35 For instance, Meisenberg therefore, this meta-analysis had significant clinical
et al36 and Berkman et al37 were able to reduce opioid heterogeneity in the modalities used for the educa-
prescriptions after developing a multimodal interven- tional interventions (paper-based, video, or both),
tion involving prescriber (ie, surgeons, nurses, and duration of the educational activity, and the topics in
anesthesiologists) and patient education. each of them, but the content of all of the interven-
Future studies may help identify which patients tions was focused on the risks associated with opioid
will benefit most from educational interventions use, pain expectations, and awareness of nonmedica-
and what additional benefits there are to educa- tion pain control methods. However, most of the stud-
tion. Identification of patients who are at higher risk ies included in this meta-analysis were performed
for long-term opioid use may theoretically allow in orthopedic populations and there was evidence
for preoperative interventions to avoid postopera- of significant statistical heterogeneity. Fourth, there
tive chronic opioid use.5 However, further evidence are external biases that may impact opioids use (law
is needed whether perioperative opioid education changes, published guidelines, and institutional poli-
among chronic opioid users may have the higher cies) and were not accounted in the statistical analysis.
impact than that in opioid-naïve population. In this Fifth, it is not possible to make final recommendations
meta-analysis, only 2 of the trials conducted sub- based on our results, as there is not a validated opioid
group analysis in chronic opioid users, demonstrating reduction threshold to consider or not perioperative
significantly higher benefit in this particular patient opioid educational initiatives. Finally, there remain
population,10,11 thus suggesting that patient selection many unanswered questions in this topic, such as the
may be key to demonstrate superior results from opi- optimal timing of education interventions (preopera-
oid education. tive versus postoperative), the benefits of such inter-
In addition to the reduction of opioid pills evi- ventions in the context of multimodal analgesia or
denced in this study, it is possible that opioid edu- Enhanced Recovery After Surgery (ERAS) programs,
cation interventions could bring further benefits in or the effects on patient-reported outcomes (PROs) at
terms of lower risk of opioid-related adverse events, long term.
drug dependency, and opioid misuse. However, these In conclusion, perioperative educational interven-
factors have not been studied as secondary outcomes tions reduce the number of opioid pills consumed
after opioid education interventions. Among other at 15 days, but there was no benefit demonstrated
targeted end points that should be considered include in terms of opioid cessation or reduction in opioid
cost outcomes of readmissions and other financial refills. Future RCTs should focus on evidence-based
costs related to drug addiction and rehabilitation. educational interventions (ie, multimedia/audiovi-
Functional outcomes could also be affected by peri- sual strategies) with strict homogeneity of material
operative education strategies, since it is known that to draw a more definitive recommendation. Further
rehabilitation programs, counseling, and behavioral evidence is also needed to determine the impact of
therapy facilitate patients returning to work.38,39 educational strategies on long-term opioid-related
There are important barriers for the implementa- outcomes (ie, misuse and addiction). E
tion of patient education interventions. Coughlin
et al40 performed a qualitative study using semis- ACKNOWLEDGMENTS
tructured interviews among essential stakeholders Individual-patient data of 2 trials was provided by
and identified 12 potential implementation barriers Dr Kristen Buono (Division of Pelvic Medicine and
including time and resource constraints, the type of Reconstructive Surgery, Urogynecology Consultants,
modality of educational material (design quality and Sacramento, CA) and Dr Jeffrey Stepan (Hospital for
packaging), and prescribers’ concerns for patient sat- Special Surgery, New York, NY).
isfaction scores (external policy and incentives). It is
therefore recommended that institutions implement DISCLOSURES
Name: Andres Zorrilla-Vaca, MD.
strategies to achieve best outcomes from periopera-
Contribution: This author helped design the study, write the
tive education interventions. manuscript, extract the data, perform statistical analysis, inter-
This meta-analysis has a number of limitations. pret the results, and approve the final manuscript.
First, the low number of studies in the analyses for Conflicts of Interest: None.

May 2022 • Volume 134 • Number 5 www.anesthesia-analgesia.org 949

Copyright © 2021 International Anesthesia Research Society. Unauthorized reproduction of this article is prohibited.
 Perioperative Opioid Education and Opioid Consumption

Name: Gabriel E. Mena, MD. 12. Darnall BD, Ziadni MS, Krishnamurthy P, et al. “My sur-
Contribution: This author helped write the manuscript, inter- gical success”: effect of a digital behavioral pain medicine
pret the results, and approve the final manuscript. intervention on time to opioid cessation after breast cancer
Conflicts of Interest: G. E. Mena receives an academic grant surgery—a pilot randomized controlled clinical trial. Pain
from PACIRA Pharmaceuticals. Med. 2019;20:2228–2237.
Name: Pedro T. Ramirez, MD. 13. Bloom DA, Baron SL, Luthringer TA, et al. Preoperative
Contribution: This author helped write the manuscript, inter- opioid education has no effect on opioid use in patients
pret the results, and approve the final manuscript. undergoing arthroscopic rotator cuff repair: a prospec-
Conflicts of Interest: None. tive, randomized clinical trial. J Am Acad Orthop Surg.
Name: Bradley H. Lee, MD. Published online December 09, 2020. DOI:10.5435/
Contribution: This author helped write the manuscript, extract JAAOS-D-20-00594.
the data, interpret the results, and approve the final manuscript. 14. Nahhas CR, Hannon CP, Yang J, Gerlinger TL, Nam D,
Conflicts of Interest: None. Della Valle CJ. Education increases disposal of unused opi-
Name: Alexandra Sideris, PhD. oids after total joint arthroplasty: a cluster-randomized con-
Contribution: This author helped write the manuscript, inter- trolled trial. J Bone Joint Surg Am. 2020;102:953–960.
pret the results, and approve the final manuscript. 15. Guyatt G, Oxman AD, Akl EA, et al. GRADE guidelines:
Conflicts of Interest: A. Sideris is supported by the C.V. Starr 1. Introduction-GRADE evidence profiles and summary of
Foundation and serves as a consultant on a cannabinoid findings tables. J Clin Epidemiol. 2011;64:383–394.
research grant funded by Colmbia Cientifica/Colciencias. 16. Wan X, Wang W, Liu J, Tong T. Estimating the sample mean
Name: Christopher L. Wu, MD. and standard deviation from the sample size, median,
Contribution: This author helped design the study, write range and/or interquartile range. BMC Med Res Methodol.
the manuscript, interpret the results, and approve the final 2014;14:135.
manuscript. 17. Egan KG, De Souza M, Muenks E, Nazir N, Korentager R.
Conflicts of Interest: C. L. Wu receives salary support through Opioid consumption following breast surgery decreases
a contract with the AHRQ (HHSP233201500020I). with a brief educational intervention: a randomized, con-
This manuscript was handled by: Tong J. Gan, MD. trolled trial. Ann Surg Oncol. 2020;27:3156–3162.
18. Lam L, Richardson MG, Zhao Z, Thampy M, Ha L,

REFERENCES Osmundson SS. Enhanced discharge counseling to reduce
1. Hah JM, Bateman BT, Ratliff J, Curtin C, Sun E. Chronic opi- outpatient opioid use after cesarean: a randomized clinical
oid use after surgery: implications for perioperative man- trial. Am J Obstet Gynecol MFM. 2020;3:100286.
agement in the face of the opioid epidemic. Anesth Analg. 19. Sabesan VJ, Chatha K, Koen S, Dawoud M, Gilot G.

2017;125:1733–1740. Innovative patient education and pain management proto-
2. Dixit AA, Chen CL, Inglis-Arkell C, Manuel SP. Assessment cols to achieve opioid-free shoulder arthroplasty. JSES Int.
of unused opioids following ambulatory surgery. Am Surg. 2020;4:362–365.
2020;86:652–658. 20. Singh S, Clarke C, Lawendy AR, Macleod M, Sanders D,
3. Bicket MC, Long JJ, Pronovost PJ, Alexander GC, Wu CL. Tieszer C. First place: a prospective, randomized controlled
Prescription opioid analgesics commonly unused after sur- trial of the impact of written discharge instructions for post-
gery: a systematic review. JAMA Surg. 2017;152:1066–1071. operative opioids on patient pain satisfaction and on mini-
4. Cozowicz C, Olson A, Poeran J, et al. Opioid prescription mizing opioid risk exposure in orthopaedic surgery. Current
levels and postoperative outcomes in orthopedic surgery. Orthopaedic Practice. 2018;29:292–296.
Pain. 2017;158:2422–2430. 21. Ilyas AM, Chapman T, Zmistowski B, Sandrowski K,

5. Soffin EM, Lee BH, Kumar KK, Wu CL. The prescription Graham J, Hammoud S. The effect of preoperative opioid
opioid crisis: role of the anaesthesiologist in reducing opi- education on opioid consumption after outpatient ortho-
oid use and misuse. Br J Anaesth. 2019;122:e198–e208. pedic surgery: a prospective randomized trial. Orthopedics.
6. Wetzel M, Hockenberry J, Raval MV. Interventions for 2021;44:123–127.
postsurgical opioid prescribing: a systematic review. JAMA 22. Buono K, Whitcomb E, Guaderrama N, et al. A random-
Surg. 2018;153:948–954. ized controlled trial assessing the impact of opioid-specific
7. Roth B, Boateng A, Berken A, Carlyle D, Vadivelu N. Post- patient counseling on opioid consumption and disposal
operative weaning of opioids after ambulatory surgery: the after reconstructive pelvic surgery. Female Pelvic Med
importance of physician stewardship. Curr Pain Headache Reconstr Surg. 2021;27:151–158.
Rep. 2018;22:40. 23. Elhage SA, Thielen ON, Huber AT, et al. Preoperative

8. Lee BH, Wu CL. Educating patients regarding pain manage- patient opioid education, standardization of prescriptions,
ment and safe opioid use after surgery: a narrative review. and their impact on overall patient satisfaction. Surgery.
Anesth Analg. 2020;130:574–581. 2021;169:655–659.
9. Kang E, Gillespie BM, Tobiano G, Chaboyer W. Discharge 24. Parsa FD, Pavlosky KK, Harbison G, et al. Effect of pre-
education delivered to general surgical patients in their operative patient education on opioid consumption and
management of recovery post discharge: a systematic well-being in breast augmentation. Plast Reconstr Surg.
mixed studies review. Int J Nurs Stud. 2018;87:1–13. 2020;145:316e–323e.
10. Syed UAM, Aleem AW, Wowkanech C, et al. Neer Award 25. Rucinski K, Cook JL. Effects of preoperative opioid edu-
2018: the effect of preoperative education on opioid con- cation on postoperative opioid use and pain manage-
sumption in patients undergoing arthroscopic rotator cuff ment in orthopaedics: a systematic review. J Orthop.
repair: a prospective, randomized clinical trial. J Shoulder 2020;20:154–159.
Elbow Surg. 2018;27:962–967. 26. Cheesman Q, DeFrance M, Stenson J, et al. The effect of
11. Stepan JG, Sacks HA, Verret CI, Wessel LE, Kumar K, Fufa preoperative education on opioid consumption in patients
DT. Standardized perioperative patient education decreases undergoing arthroscopic rotator cuff repair: a prospective,
opioid use after hand surgery: a randomized controlled randomized clinical trial-2-year follow-up. J Shoulder Elbow
trial. Plast Reconstr Surg. 2021;147:409–418. Surg. 2020;29:1743–1750.

950   
www.anesthesia-analgesia.org ANESTHESIA & ANALGESIA
Copyright © 2021 International Anesthesia Research Society. Unauthorized reproduction of this article is prohibited.
 E  Meta-Analysis

27. Kwan SY, Lancaster E, Dixit A, et al. Reducing opioid use in 34. Hillman RT, Iniesta MD, Shi Q, et al. Longitudinal patient-
endocrine surgery through patient education and provider reported outcomes and restrictive opioid prescribing after
prescribing patterns. J Surg Res. 2020;256:303–310. minimally invasive gynecologic surgery. Int J Gynecol
28. Hite M, Dippre A, Heldreth A, et al. A multifaceted
Cancer. 2021;31:114–121.
approach to opioid education, prescribing, and disposal for 35. Yorkgitis BK, Paffett C, Brat GA, Crandall M. Effect of sur-
patients with breast cancer undergoing surgery. J Surg Res. gery-specific opioid-prescribing education in a safety-net
2021;257:597–604. hospital. J Surg Res. 2019;243:71–74.
29. Derefinko KJ, Salgado García FI, Johnson KC, et al. A ran- 36. Meisenberg BR, Grover J, Campbell C, Korpon D.

domized pilot program to reduce opioid use following Assessment of opioid prescribing practices before and
dental surgery and increase safe medication return. Addict after implementation of a health system intervention
Behav. 2020;102:106190. to reduce opioid overprescribing. JAMA Netw Open.
30. Farley KX, Anastasio AT, Kumar A, Premkumar A,
2018;1:e182908.
Gottschalk MB, Xerogeanes J. Association between quan- 37. Berkman RA, Wright AH, Sivaganesan A. Opioid-free spine
tity of opioids prescribed after surgery or preoperative surgery: a prospective study of 244 consecutive cases by a
opioid use education with opioid consumption. JAMA. single surgeon. Spine J. 2020;20:1176–1183.
2019;321:2465–2467. 38. Archer KR, Devin CJ, Vanston SW, et al. Cognitive-

31. Zhang DDQ, Sussman J, Dossa F, et al. A systematic review behavioral-based physical therapy for patients with chronic
of behavioral interventions to decrease opioid prescribing pain undergoing lumbar spine surgery: a randomized con-
after surgery. Ann Surg. 2020;271:266–278. trolled trial. J Pain. 2016;17:76–89.
32. Mears SC, Shnaekel A, Wilkinson J, Chen C, Barnes CL. A 39. Coronado RA, Ehde DM, Pennings JS, et al. Psychosocial
departmental policy can reduce opioid prescribing after mechanisms of cognitive-behavioral-based physical ther-
orthopedic surgery. J Opioid Manag. 2019;16:41–47. apy outcomes after spine surgery: preliminary findings
33. Lovecchio F, Stepan JG, Premkumar A, et al. An insti- from mediation analyses. Phys Ther. 2020;100:1793–1804.
tutional intervention to modify opioid prescribing 40. Coughlin JM, Shallcross ML, Schäfer WLA, et al. Minimizing
practices after lumbar spine surgery. J Neurosurg Spine. opioid prescribing in surgery (MOPiS) initiative: an analy-
2019;30:483–490. sis of implementation barriers. J Surg Res. 2019;239:309–319.

May 2022 • Volume 134 • Number 5 www.anesthesia-analgesia.org 951

Copyright © 2021 International Anesthesia Research Society. Unauthorized reproduction of this article is prohibited.