Pharmacists' attitudes, knowledge, utilization, and outcomes involving prescription drug monitoring programs: A brief scoping review

SCIENCE AND PRACTICE Journal of the American Pharmacists Association xxx (2018) 1e9

Contents lists available at ScienceDirect

Journal of the American Pharmacists Association journal homepage: www.japha.org

REVIEW

Pharmacists' attitudes, knowledge, utilization, and outcomes involving prescription drug monitoring programs: A brief scoping review Kirbee Johnston, Lindsey Alley, Kevin Novak, Sarah Haverly, Adriane Irwin, Daniel Hartung* a r t i c l e i n f o

a b s t r a c t

Article history: Received 9 February 2018 Accepted 3 June 2018

Objective: While literature on pharmacists’ engagement with prescription drug monitoring programs (PDMPs) is growing, no formal synthesis of findings has been conducted to provide overarching recommendations for research or practice. The objective of this study was to identify and synthesize findings from current literature on community pharmacists’ attitudes toward, knowledge of, and registration and utilization behaviors regarding PDMPs. Data sources: Electronic databases (MEDLINE, PsychINFO, Cochrane Database of Systematic Reviews, Google Scholar, and the Brandeis University PDMP Center of Excellence) and reference lists from relevant manuscripts were searched for relevant English-language manuscripts. Key words used in searches included pharmacist, prescription drug monitoring program, opioid safety, attitudes, knowledge, and utilization. Study selection: Papers were included from January 1, 2008 up to October 6, 2017. Three authors independently screened articles for full text review; 2 authors independently conducted full text review for final study selection. Discrepancies were resolved through consensus. Data extraction: Data were extracted to an evidence table, coded by topic category, and checked for accuracy. Results: Fifteen manuscripts met inclusion criteria. The studies varied greatly in methodological approach. In general, pharmacists’ attitudes and knowledge of PDMPs positively influenced likelihood to register and use their state’s program. Targeted training had a substantial impact on knowledge, registration, and utilization. Conclusion: Pharmacist-targeted PDMPs and opioid safety training is highly recommended to increase knowledge of and insight into behavioral change. © 2018 American Pharmacists Association®. Published by Elsevier Inc. All rights reserved.

As the opioid crisis continues to escalate in the United States, federal, local, and academic efforts are under way to examine opioid-related harms and their relationship to prescribing practices.1,2 While overdose from heroin and illicitly

Disclosure: The authors declare no conflicts of interest or financial interests in any product or service mentioned in this article. Funding: Funding for this project was provided by the Agency for Healthcare Research and Quality (AHRQ 5R18HS024227-02). Previous presentation: Some of the concepts and data appearing in this manuscript have been presented at the Rx Opioid and Heroin Summit, Atlanta, GA (April 17-20, 2017) and the American Public Health Association Conference, Atlanta, GA (November 4-8, 2017). * Correspondence: Daniel Hartung, PharmD, MPH, Oregon State UniversityeOregon Health and Science University College of Pharmacy, 2730 SW Moody Ave., Mailcode: CL5CP, Portland, OR 97201. E-mail address: [email protected] (D. Hartung).

manufactured synthetic opioids constitutes more than half of all opioid overdoses, prescription opioids contributed to more than 17,000 deaths in 2016, an increase of 10.6% from 2015.3 In the same year, over 200 million opioid prescriptions were dispensed in pharmacies across the United States, at a rate of 66.5 per 100 persons.4 Current efforts to reduce inappropriate opioid prescribing include prescriber education, medication surveillance, and health care system and payer policy restrictions. Although most interventions have been directed at prescribers, pharmacists play an important role in ensuring patient safety during the treatment process, and they have arguably received the least attention in opioid safety and outcomesrelated research. The community pharmacy profession has evolved considerably to embrace more patient-centered medication management models of care, and pharmacists’

https://doi.org/10.1016/j.japh.2018.06.003 1544-3191/© 2018 American Pharmacists Association®. Published by Elsevier Inc. All rights reserved.

SCIENCE AND PRACTICE K. Johnston et al. / Journal of the American Pharmacists Association xxx (2018) 1e9

Key Points Background:  The role of community pharmacy in addressing the opioid crisis is expanding rapidly, but opportunities for pharmacist intervention remain ill defined; tailored trainings and policies are inconsistent or unavailable.  Prescription drug monitoring programs (PDMPs) provide community pharmacists an opportunity to minimize opioid-related harms and improve patient safety.  Understanding pharmacists’ existing attitudes, knowledge, and utilization regarding the PDMP is critical to improve future intervention research and enhance PDMP use. Findings:  Attitudes regarding the PDMP and knowledge of its use positively influenced pharmacists’ likelihood to register and use the program.  Pharmacist-targeted training on PDMP use and importance greatly affected knowledge, registration, and utilization.  Methodologically rigorous, theoretically grounded studies are needed to measure these effects more thoroughly and accurately.

roles in opioid safety are expanding quickly, although best practices and expectations remain ill defined. Prescription drug monitoring programs (PDMPs) provide a potential avenue for pharmacists to better engage patients and providers around opioid safety by facilitating enhanced patient counseling, monitoring for opioid safety risks, potentially intervening to prevent misuse or abuse, and recommending naloxone if appropriate.5-8 PDMPs are state-specific electronic databases that contain regularly updated patient data for scheduled drugs, and they allow physicians and pharmacists to see patterns of controlled substance prescribing and dispensing through patient reports.6,7 PDMPs are currently active in 49 states, with varying mandated-use and registration policies. Despite the potential benefits to improve patient safety, many health care providers report difficulty integrating PDMP queries into their daily workflow.6,9-11 Commonly stated barriers include unfamiliarity with or dislike of the user interface, difficulties working collaboratively with providers or patients in response to troubling PDMP reports, concerns about legal liability, and unwillingness to devote time regularly engaging the program when practicing in a state without a mandated use policy.6,10,11 All these barriers could be addressed through targeted and tailored training, as it is established that training plays an important role in shaping attitudes and knowledge.12,13 It is critical that investigators begin objectively and comprehensively examining the relationship pharmacists

2

have with their state’s PDMP. The purpose of the current review is to identify and synthesize literature specific to community pharmacists’ attitudes and knowledge regarding PDMP use and communication, behavioral intentions to use the program, and actual registration, utilization, and outcomes. Specifically, the following areas were examined: 1) community pharmacists’ knowledge and attitudes regarding PDMP registration, use, and impact; 2) the impact of attitudes and knowledge on PDMP registration and utilization; and 3) dispensing and patient health outcomes resulting from pharmacists’ use of the PDMP. An overview of current findings and recommendations for future research areas and methodologies are provided. Methods A scoping review was conducted to examine the areas of interest. Scoping reviews summarize the breadth and depth of current literature to address broad topics.14,15 Scoping reviews are beneficial when the literature uses varying methodologies, no prior review has been performed to synthesize the topic, and identifying a narrow research question is difficult.16 Unlike systematic reviews, scoping reviews do not aim to evaluate the quality of the literature.16 An extensive search was conducted of MEDLINE, PsychINFO, Cochrane Database of Systematic Reviews, Google Scholar, the Brandeis University PDMP Center of Excellence, and reference lists from relevant manuscripts. The search strategy used combinations of the following terms: pharmacy or pharmacist AND prescription monitoring program, prescription drug monitoring program, controlled substance database, PDMP, PMP, attitudes, knowledge, opioids, opioid safety, dispensing, or utilization. In addition, a call was placed for unpublished manuscripts and data to several pharmacy-based listservs (Figure 1). The eligible publication dates used in the search ranged between January 1, 2008 and October 6, 2017. To be included, articles needed to be written in English, include community pharmacists in the study sample, report on PDMPs operating in the United States, and address at least one of the areas of interest: attitudes, knowledge, registration, and utilization. Studies were screened independently and coded for full-text review by 3 of the authors. Full-text review was also conducted independently for final study selection by 2 authors. Discrepancies were resolved through consensus, although agreement on “articles for inclusion” was high (Cohen’s k ¼ 0.91). Data were extracted into an evidence table and checked for accuracy. Methodologies, including measurement tools and the operationalization of outcome variables, varied considerably across studies, making a quantitative meta-analysis unfeasible. The few studies examining outcomes of community pharmacists’ PDMP use defined outcomes in terms of dispensing practices and pharmacist-patient education and communication. Thus, a qualitative synthesis was performed. Findings should be considered in the context in which they were reported, as PDMP-related attitudes and knowledge may be greatly influenced by the state and year in which the data were collected.

SCIENCE AND PRACTICE Pharmacy PDMP scoping review

Figure 1. Preferred reporting items for systematic reviews and meta-analyses (PRISMA) flow diagram.

Results Fifteen articles met all inclusion criteria and were retained for analysis. Publication years ranged from 2010 to 2016. Although inclusion was not restricted by methodology, all included studies reported survey-based results (n ¼ 13 crosssectional; n ¼ 2 pre-post design), representing 7396 licensed community pharmacist participants. One publication also used PDMP data to describe the trajectory of provider registration (Table 1). Attitudes toward the PDMP Eight studies examined pharmacists’ attitudes and perceptions of PDMPs and PDMP use. In addition to general attitudes toward PDMPs, the definition of attitudes in these studies also included perceptions and beliefs regarding effectiveness and usefulness of the information recorded in the PDMP and the pharmacists’ role in addressing the opioid crisis. Three studies exploring pharmacists’ general attitudes toward the PDMP found that pharmacists in Maine, Virginia, and Texas had overall positive views or attitudes regarding the PDMP and PDMP use.17-19 The studies that included attitudes toward convenience found that between 49% and 69% of pharmacists believed the PDMP to be convenient.15,19 Four studies in this review explored the perceptions and specific attitudes toward PDMPs’ usefulness in

combating the opioid crisis; they found that pharmacists generally believed PDMPs to be effective.20-23 Two of these studies found that 63%-81%20 of Florida pharmacists and 90% 23 of Kansas pharmacists believed PDMPs to be effective in reducing doctor shopping. A study also found that 92% 23 of pharmacists believed that PDMPs could be effective in preventing abuse and diversion, and an additional study found that this belief increased following education.21 Only one study examined pharmacists’ attitudes toward their roles in preventing opioid abuse and how that affects PDMP use. This study found that only 43% of pharmacists perceived themselves as “health care providers” in most situations with respect to opioid safety. Fifty-three percent of pharmacists believed that they were “sometimes like a health care provider, sometimes like the police.”24 The authors of this study observed that pharmacists who perceived themselves more as health care providers may be more likely to use the PDMP. Two studies used the Theory of Planned Behavior32 to study the relationship of pharmacists’ attitudes and perceived behavioral control on PDMP engagement. These studies found that pharmacists with high attitude or high perceived behavioral control ratings were twice as likely to report high intention to use PDMP.18,19 Perceived behavioral control is defined as the level of ease or difficulty with which a person believes a behavior can be performed.

3

Citation Piper et al., 201617

Gavaza et al., 201418

Attitudes Knowledge Registration

Utilization

Setting: Maine N ¼ 275 (21.8% response rate) Female ¼ 48.7% Years practice ¼ 41% with >25 years PharmD ¼ 40.5%

X

X

A majority of respondents had favorable views of PDMP (75.3%), while only more than half use it (56.2%). PharmD-educated pharmacists had a slightly more favorable rating of the PDMP than those with a bachelor's degree (P ¼ 0.055) and are significantly more likely to use PDMP (P < 0.05).

X

Pharmacists in general had a strong and positive intention to use a PDMP (mean ¼ 5.3 ± 4.6; possible range: e9 to 9) and had a moderate to positive attitude toward using PDMP (mean ¼ 6.3 ± 5.3; possible range: e12 to 12). When taking all 3 TPB model constructs into account in regression model, attitude (P < 0.001) significantly predicted pharmacists’ utilization intent.

X

When assessing pharmacists’ attitudes toward PDMP utilization, most saw it as useful (70.5%), beneficial (70.0%), good (69.6%), and valuable (69.2%), while less than half (48.5%) saw utilization as convenient. There was an overall strong positive attitude toward PDMP utilization (mean score, 11.4 ± 5.3; range, e15 to 15). Adjusted for other TPB constructs, for every one unit increase in the score for attitude, the odds of high intention of PDMP utilization increased almost 2-fold (OR, 1.8; 95% CI 1.2-2.8). The majority self-reported their PDMP knowledge as fair (33.1%; good ¼ 26.8%; very good ¼ 13.3%). The majority across practice settings believed that a PDMP could decrease the incidence of doctor shopping (63%-81%). Most disagreed that the PDMP would discourage them from dispensing controlled substances (50%-64%). All participants were aware of the PDMP, and nearly all (95.8%) were already registered in the PDMP. A large majority of pharmacists (91.3%) have used the PDMP at least once, with 62.5% reporting they use the PDMP more than once a week. The majority of participants, post-education, reported that the PDMP was a program for physicians and pharmacists to help prevent abuse and diversion (12.5% pre to 73.9% post) and that the PDMP is a public health tool (0% pre to 21.7% post). Posteducation, pharmacists indicated that they

Fass and Hardigan, 201120

To assess Florida pharmacists’ attitudes toward Setting: Florida implementing a PDMP in the state N ¼ 911 (18.2% response rate) Mean years practice ¼ 23 ± 15 years

X

X

Fleming et al., 201621

Setting: Texas To provide education to community pharmacists regarding the registration and use N ¼ 24 of the Texas PDMP and to assess the impact of the education on pharmacists’ perceptions of the PDMP

X

X

X

X

K. Johnston et al. / Journal of the American Pharmacists Association xxx (2018) 1e9

Fleming et al., 201419

Sample characteristics

To describe practicing pharmacists' beliefs regarding the misuse and diversion of opioids and their use of the PDMP; to characterize the prevalence of opioid use relative to other controlled substance use according to the PDMP; and to evaluate patterns of arrests involving opioids, with an emphasis on Schedule II prescription agents To determine what Theory of Planned Behavior model constructs (attitude, subjective norm, perceived behavioral control), past utilization behavior, and perceived moral obligation were significant predictors of pharmacists’ intention to use the PDMP

X Setting: Virginia N ¼ 97 (16.2% response rate) Mean age ¼ 49.5 ± 13.4 years Female ¼ 49.5% White/non-Hispanic ¼ 96.4% Mean years practice ¼ 23.2 ± 13.4 years X Setting: Texas To explore the Theory of Planned Behavior’s N ¼ 261 (26.2% utility in predicting Texas pharmacists' intention to use an online accessible PDMP; to response rate) determine the contribution of each construct; Mean age ¼ 50.3 ± 13.7 and to text whether the addition of perceived years Female ¼ 47.9% obligation was significantly related to White/non-Hispanic ¼ pharmacists’ intention 65.8% Mean years practice ¼ 22.3 ± 14.4 years PharmD ¼ 29.2%

Outcomes

Main findings

Objectives

X

SCIENCE AND PRACTICE

4

Table 1 Evidence table: Characteristics and findings of each study meeting inclusion criteria

X

Wixson et al., 201523 To identify characteristics of Kentucky community pharmacists and their practice environment associated with utilization of Kentucky’s PDMP KASPER

Setting: Kentucky N ¼ 563 (27.9% response rate) Median years practice by practice type (IQR): Independent ¼ 16 years (35 years) Chain ¼ 22 years (60 years)

X

X

Fendrich et al., 201524

To assess the attitudes of pharmacists during encounters with patients suspected of misusing prescriptions; to assess pharmacist’s PDMP utilization 1 year after implementation

Setting: Midwestern state N ¼ 48 (29.6% response rate)

X

Fleming et al., 201325

To describe PDMP utilization by prescribers, pharmacists, and law enforcement for active state PDMPs; PDMP utilization by health care providers with and without online access; average annual operational costs for PDMPs from 2008 to 2009; and PDMP requests based on PDMP housing authority (law enforcement vs. nonelaw enforcement) To describe outreach efforts employed to encourage use of Oregon’s relatively new

Setting: multiple states N ¼ 15 (45.5% response rate) PDMP administrators

Deyo et al., 201426

X

X

X

X

X

X

More than 98% of pharmacies had submitted data to PDMP in 2012. There were 7 million (continued on next page)

5

SCIENCE AND PRACTICE

Setting: Ohio X N ¼ 1434 (25% response rate) Mean years practice: Enrollees ¼ 17.4 ± 10.4 years Nonenrollees ¼ 15.6 ± 10.6 years

Pharmacy PDMP scoping review

Ulbrich et al., 201022 To determine factors influencing enrollment for community pharmacists registered and not registered to Ohio’s PDMP OARSS; to identify association of OARRS with demographics, availability of Internet access at work, educational background, and/or previous PDMP education received; and to compare knowledge of OARRS in enrollees versus nonenrollees

would be very likely to use the Texas PDMP within the next 30 days (73.7%). More than half of community pharmacists were registered in OARRS (62%). Nonregistered pharmacists stated the top factor influencing registration was “time available to access OARRS report.” Registered pharmacists stated the top factor influencing registration was “being able to assist with decreasing doctor shopping.” When asked to identify medications in the OARRS database, registered pharmacists correctly identified significantly more medications than nonregistered pharmacists (P < 0.001). The majority of respondents reported either they or personnel in their pharmacy have used KASPER since its inception (84%). Respondents viewed KASPER as an effective tool to decrease drug abuse and diversion (92%) and doctor shopping (90%). Adjusted for confounding, the odds of using KASPER among community pharmacists in urban locations were 1.11 times higher than the odds among community pharmacists in rural locations (95% CI 1.011.21), and the odds of using KASPER among independent pharmacists were 1.27 times higher than the odds among chain pharmacists (95% CI 1.15-1.40). When pharmacists were asked about their role regarding the dispensing of scheduled prescription medications, 43% saw their role in most situations as a “health care provider,” while 53% saw their role in most situations as “sometimes like a health care provider, sometimes like the police.” Pharmacists with a “health care provider” view were more likely to always counsel patients suspected of misusing prescriptions about risk (55% vs. 27%; P < 0.10) and to always advise patients about addiction treatment (20% vs. 4%, P < 0.10). Regarding pharmacists’ use of the PDMP, most found it easy to input information (66%) and retrieve information (80%). The aggregate rate of pharmacists querying the PDMP was 268.9 per 100,000 population (SD ¼ 261.2). The aggregate query rate was lower in states where PDMPs were under the authority of law enforcement (119 [SD ¼ 158] vs. 295 [SD ¼ 261]). On average across states, 20.9% of pharmacists were registered in the PDMP.

Citation

Objectives

Sample characteristics

Attitudes Knowledge Registration

Utilization

Outcomes

Setting: Oregon PDMP; examine the pace of registration and utilization of the program; identify barriers to N ¼ All PDMPregistered providers registration and use; and identify system improvements that might enhance the value of such programs

To examine how integration and consistent use of a PDMP in pharmacy practice impacts pharmacists’ dispensing practices related to CSPs

Setting: Indiana N ¼ 1582 (15% response rate) Mean age ¼ 46.9 ± 13.7 years Female ¼ 54% White/non-Hispanic ¼ 93% Mean years practice ¼ 20.8 ± 14 years

X

Norwood and Wright, 201628

Setting: Indiana N ¼ 1000 sample taken from 1582 who complete survey (15% response rate) Mean age ¼ 46.4 ± 13.6 years Female ¼ 54% White/Non-Hispanic ¼ 93% Mean years practice ¼ 20.5 ± 14 years To assess differences in PDMP use between 2 Setting: Connecticut adjacent states with different PDMP pharmacist and Rhode Island N ¼ 294 (198 in CT, 96 accessibility, to explore use of PDMPs in pharmacy practice, and to examine associations in RI; 10% response between PDMP use and pharmacists’ responses rate) Age 45-60 years ¼ 41% to suspected diversion or doctor shopping (CT) and 45.3% (RI) Female ¼ 55.6% (CT) and 50% (RI) >10 years licensed ¼ 75.9% (CT) and 81.1% (RI)

X

Green et al., 201329

To identify barriers to PDMP use in outpatient pharmacies and determine the impact these barriers have on utilization

X

X

X

Main findings prescriptions entered in 2012. When examining percentage of PDMP accounts by specialty, pharmacists comprise 22% of accounts. System queries increased in the second quarter of 2013 because of pharmacists using the system more often. Two large pharmacy chains introduced policies that required pharmacists to register with the PDMP and to query it before dispensing certain controlled substances. This policy change in the 2 chains inspired pharmacists from other chains to register. Ninety-four percent of respondents had heard of Indiana’s PDMP (INSPECT), but only 72% of respondents reported using INSPECT. Those who felt INSPECT provided increased access to patient information were 6.4 times more likely to dispense fewer controlled substance prescriptions. Pharmacists who always use INSPECT refused to dispense more CSPs per year than those who never use INSPECT (mean ¼ 24.8 ± 27.3 vs. 6.9 ± 13.1). Furthermore, these always-users were 3.3 times more likely to refuse to dispense more CSPs than never-users. Nearly all outpatient pharmacists (97%) were aware of Indiana’s PDMP (INSPECT), while only 81% have used it. Those extremely concerned with prescription drug abuse in community were significantly more likely than those not concerned to both use PDMP (OR, 9.96; 95% CI 1.72-57.54) and frequently use PDMP (OR, 17.89; 95% CI 1.46-219.69).

Use of PDMP was higher in CT than in RI (67.9% and 7.8%, respectively). When faced with suspicious medication use behavior, pharmacists who used the PDMP were less likely than nonusers to tell the patient they were out of stock of the drug (aOR, 0.27; 95% CI 0.12-0.60) and to discuss their concerns with the patient (aOR, 0.48; 95% CI 0.25-0.92). There was no significant difference between users and nonusers in the odds of contacting the patient’s provider (aOR, 0.86; 95% CI 0.213.47), referring the patient back to the prescriber (aOR, 1.50; 95% CI 0.79-2.86), or refusing to fill the prescription (aOR, 0.63; 95% CI 0.79-2.86).

K. Johnston et al. / Journal of the American Pharmacists Association xxx (2018) 1e9

Norwood and Wright, 201527

SCIENCE AND PRACTICE

6

Table 1 (continued)

SCIENCE AND PRACTICE

Abbreviations used: OR, odds ratio; CI, confidence interval; PDMP, prescription drug monitoring program; OARRS, Ohio Automated Rx Reporting System; CSP, controlled substance prescription; INSPECT, Indiana Scheduled Prescription Electronic Collection and Tracking; aOR, adjusted odds ratio; CS, controlled substance; CSP, controled substance prescription; CPE, continuing pharmacy education; TPB, Theory of Planned Behavior; KASPER, Kentucky All Schedule Prescription Electronic Reporting.

A majority of pharmacists who frequently use PDMP state that they use it when they have any suspicion of misuse or drug diversion (92%). Among pharmacists who use the PDMP, 54% have decreased number of CS prescriptions dispensed, and 45% provided more patient education or counseling on CS use and abuse. X Rittenhouse et al., 201531

Fleming et al., 201430

To examine situations that would prompt pharmacists to access a patient’s controlled substance history using an online-accessible PDMP database and to assess pharmacists’ views on managing opioid addiction. The study also explored the actions taken by pharmacists who suspect abuse, the tasks associated with dispensing CSPs, and CPE training related to pain management and diversion.

Setting: Texas N ¼ 261 (26.2% response rate) Mean age ¼ 50.3 ± 13.7 years Female ¼ 47.9% White/non-Hispanic ¼ 65.8% Mean years practice ¼ 22.3 ± 14.4 years PharmD ¼ 29.1% Setting: Arkansas To determine the makeup of Arkansas PDMP N ¼ 631 (41.7% users; to describe utilization of PMP; to response rate for examine the performance of the program in regard to changing prescribing and dispensing overall sample including other medical practices professions)

X

Pharmacists stated situations that would always trigger PDMP use include mistakes or irregularities in prescriptions (68%), early refill requests (66%), and patients who pay in cash (48%). When responding to a suspected abuse, the majority of pharmacists agreed or strongly agreed with calling the prescriber (92%), refusing to dispense the prescription (90%), and documenting the incident (79%). X

Pharmacy PDMP scoping review

PDMP knowledge and training Three studies examined pharmacists’ self-reported knowledge of PDMPs, and 2 studies examined pharmacist training in using PDMPs (employing pretests and posttests to gauge learning). One study in Florida found that more than half of participating pharmacists self-reported having little to no knowledge of the PDMP, whereas only 5.4% reported having excellent knowledge of it.20 Interestingly, another study reported that most pharmacists in a midwestern state found it easy to input (66%) and to extract (80%) information from the PDMP.24 A study examining enrollment into Ohio’s PDMP found that 65.7% of enrolled community pharmacists believed that knowledge of the PDMP was very or somewhat important for intending to enroll, and 62.2% believed that experience with the PDMP at their work setting was very or somewhat important for their own enrollment.22 One study in Texas examined the impact of a pharmacist-targeted PDMP training and education intervention. It found that the percent of pharmacists who defined the purpose of the PDMP as a program for preventing abuse and diversion went from 12.5% before education to 73.9% after education.21 PDMP registration Four articles explored issues surrounding pharmacistreported registration and actual registration rates in their respective states’ PDMP. The aforementioned investigation of PDMP registration in Ohio found disparity in enrollment motivations across the 2 groups, such that nonenrollees believed that a lack of time and pharmacy Internet access were the primary factors influencing enrollment, whereas enrollees believed that the ability to decrease doctor shopping and drug diversion were the primary factors.22 Pharmacy managers were more likely to be enrolled in Ohio’s PDMP, as were pharmacists with greater years in practice, pharmacy Internet accessibility, and some level of PDMP education or outreach (i.e., mailed brochures). Similarly, a pilot educational intervention in Texas reported that although the majority of pharmacist participants were registered to use the PDMP, they lacked an understanding of how the program could be useful in their setting, and the few nonregistered participants reported that lack of time and knowledge were the primary barriers to registration.21 A multistate survey of PDMP administrators across 33 nonmandated-use states revealed that only 20.9% of pharmacists registered for their program as compared with 53.2% of physicians.25 An epidemiologic investigation of Oregon’s PDMP observed that pharmacy chain policies had a significant impact on pharmacist PDMP registration and use.26 This observation suggests that store- or practice-level policies mandating provider registration and use may bolster program registration and engagement, even in states like Oregon, where mandated use policies are absent. PDMP use and outcomes Twelve articles examined use and outcomes related to dispensing practices and pharmacist-patient education and communication. In 5 papers spanning 4 years, the percentage of pharmacists who reported using their PDMP ranged from 7.8% in Rhode Island (2013) to 91.3% in Indiana (2016).17,21,27-29 A study 7

SCIENCE AND PRACTICE K. Johnston et al. / Journal of the American Pharmacists Association xxx (2018) 1e9

in Kentucky found that 84% of responding pharmacists reported that they or personnel in their pharmacy had used the state PDMP since its inception.23 Two studies provided information on pharmacists’ PDMP query rates.25,26 When examining the rate of Oregon pharmacists’ PDMP queries over time, a noticeable increase in the rate was found to be driven by mandatory query policies implemented by large pharmacy chains.26 In 2 studies examining pharmacist intent, researchers found that pharmacists generally had a strong intention to use their PDMP and that having a positive attitude regarding PDMPs was significantly associated with intention to use the PDMP.18,19 One study reported a significantly higher likelihood of using the PDMP if the pharmacist worked at an independent pharmacy or their pharmacy was in an urban area.23 Stated prompts for using the PDMP included commonly cited “red flags,” such as patients paying in cash,30 mistakes or irregularities in the prescription,30 early refill requests,30 and any suspicion of misuse or drug diversion.31 Three manuscripts explored the outcomes of pharmacists’ PDMP use. Self-reported use generally led to pharmacists dispensing fewer controlled substance prescriptions,27,31 but not always.29 Other outcomes included providing more patient education and counseling on controlled substance use and abuse31 and being less likely than nonusers to tell patients that they were out of stock of the medication.29

Discussion The studies included in this review reveal that PDMP-related behavior change is often motivated through pharmacists’ attitudes and beliefs about the behavior, their knowledge regarding how to perform the behavior (e.g., register or use the PDMP), and their intention to act. The major challenge pharmacists appear to face when dispensing opioids is maintaining a balance between the gatekeeper role (e.g., protecting patients and the public from opioid misuse) and the caregiver role (e.g., providing consultation and care).7,33 Pharmacists also reported having suboptimal collaboration and communication with physicians, lacking education and training for discussions on substance use disorders opioid safety communication, having limited time and reimbursement for clinical time with patients, and having fears of legal and ethical ramifications.33-35 Pharmacists who were provided training or who perceived that they had already been trained effectively reported significantly more favorable attitudes toward the PDMP, higher PDMP registration rates, and improved opioid safety knowledge. The provision of pharmacist-targeted training that included information about opioid safety risks relevant to the pharmacists’ own community further significantly improved registration and utilization rates. PDMP functionality and user interface vary significantly between states, and they have evolved significantly over time. In the span of only a few years, the PDMP data lag for new controlled substance fills has reduced from 1 month to 1 week to 72 hours and, in some states, prescription data are uploaded daily; in addition, a growing number of PDMPs now allow for interstate data sharing.36,37 In many states, the standard PDMP interface has been enhanced to include provider alerts, pharmacy or clinic maps, risk scores, milligram morphine equivalent graphs, and medication assistant treatment referral notes.8,38,39 Synthesizing literature across multiple states spanning a decade to compile overarching themes and recommendations was, thus, inherently difficult. Barring the 8

implementation of nationwide PDMP, mandated use policies, or standard interface designs, the interstate evaluation of findings will remain difficult. Current research on PDMP functionality is under way to evaluate the various PDMP interface enhancements in multiple states.40,41 The results reported here should be interpreted with caution and an appreciation for the context in which the data were collected. Limitations and recommendations This scoping review was hindered by an overall lack of pharmacist-focused literature on PDMP-related attitudes, knowledge, registration, and utilization. The number of empirical, epidemiologic, and intervention studies focusing on the relationships between pharmacists and opioid risk prevention in general are severely limited. Among the included articles, is a general inconsistency in methodologies, including measurements tools and the operationalization of outcome variables that made quantitative synthesis unfeasible. Only 2 of the included articles were theoretically grounded, using the Theory of Planned Behavior32 to develop and validate survey instruments.18,19 Although pharmacists are increasingly recognized as key stakeholders in the fight against the opioid crisisdespecially given the rapid implementation of pharmacy-based naloxone legislation across statesdthe evidence to support training and intervention development is sorely lacking.42,43 As this area of research continues to grow, future scoping and systematic reviews should continue attempting to consolidate and disseminate overarching lessons from which to build public health and administrative interventions. Those who are working to disseminate research in this area are often impaired by a lack of validated measurement tools, theoretical foundations, and small sample sizes. Fleming et al. are among a small number of pharmaceutical health services researchers to publish measures specific to pharmacists’ role in addressing opioid safety18,19,29,30,35 and the only to have validated their instruments.19 Building on their success, future research in this area should strive to develop and use validated measures from which substantive evidence and conclusions may be drawn and upon which replication studies can be performed accurately to support the science further. These studies should build on established theories and others’ empirical findings. Conclusion The provision of education is an important form of intervention for affecting behavioral change in pharmacists. Research has established that targeted training is one of the most effective mechanisms for improving workers’ skillsets and inciting behavioral changes, above and beyond the impact of appraisal and feedback, management by objectives, and other psychological and behavioral interventions.12 It is important for future work to examine how PDMP training specifically increases pharmacists’ attitudes and knowledge around opioid safety and patient engagement and how it can directly improve patient overdose risk. Pharmacy managers and medical directors should look to training as the primary means of preparing staff members, including management, to address the opioid crisis. With tools like the PDMP and knowledge and skills gained from training, it is possible to increase the impact that pharmacists have on maintaining patient safety surrounding the dispensing of prescription

SCIENCE AND PRACTICE Pharmacy PDMP scoping review

opioids. Finally, it will be important to evaluate the effects of increased pharmacist engagement in this area on patient outcomes. Acknowledgments Kirbee Johnston, Lindsey Alley, and Kevin Novak contributed equally to this work and are considered co-first authors. Adriane Irwin and Daniel Hartung contributed equally to this work. References 1. Nayak R. The opioid crisis is now full-blown! Are pharmacists ready? J Pharm Prac Edu. 2018;1(1):1. 2. Wetter S, Hodge JG, Chronister D, Hess A. The opioid epidemic and the future of public health emergencies. ABA Health eSource. 2018;14(5). Available at: https://papers.ssrn.com/sol3/papers.cfm?abstract_id¼3112 908. Accessed July 11, 2018. 3. Seth P, Scholl L, Rudd RA, Bacon S. Overdose deaths involving opioids, cocaine, and psychostimulantsdUnited States, 2015e2016. MMWR Morb Mortal Wkly Rep. 2018;67:349e358. 4. Centers for Disease Control and Prevention. U.S. prescribing rate maps. Atlanta, GA: CDC; 2017. Available at: https://www.cdc.gov/ drugoverdose/maps/rxrate-maps.html. Accessed July 11, 2018. 5. Joranson DE, Elliot D, Lipman AG. Pain and the pharmacist. Pain Med. 2003;4(2):190e194. 6. Hoppe J, Howland MA, Nelson L. The role of pharmacies and pharmacists in managing controlled substance dispensing. Pain Med. 2014;15(12): 1996e1998. 7. American Pharmacists Association. Pharmacists’ role in addressing opioid abuse, addiction, and diversion. J Am Pharm Assoc. 2014;54(1):e5ee15. 8. Weiner J, Bao Y, Meisel Z. Prescription drug monitoring programs: evolution and evidence. LDI Issue Briefs. 2017. Available at: https://ldi.upenn. edu/brief/prescription-drug-monitoring-programs-evolution-and-evidence. Accessed July 11, 2018. 9. Hildebran C, Cohen DJ, Irvine JM, Foley C, O'kane N, Beran T, et al. How clinicians use prescription drug monitoring programs: a qualitative inquiry. Pain Med. 2014;15(7):1179e1186. 10. Hartung DM, Hall J, Haverly SN, Cameron D, Alley L, Hildebran C, et al. Pharmacists’ role in opioid safety: a focus group investigation. Pain Med. 2017:pnx139. Available at: https://doi.org/10.1093/pm/pnx139. Accessed July 11, 2017. 11. Barlas S. Pharmacists step up efforts to combat opioid abuse: the CDC and congress are trying to pitch in. P T. 2015;40(6):369e401. 12. Arthur Jr W, Bennett Jr W, Edens PS, Bell ST. Effectiveness of training in organizations: a meta-analysis of design and evaluation features. J Appl Psychol. 2003;88(2):234e245. 13. Pratt SD, Mann S, Salisbury M, Greenberg P, Marcus R, Stabile B, et al. Impact of CRMebased team training on obstetric outcomes and clinicians’ patient safety attitudes. Jt Comm J Qual Patient Saf. 2007;33(12):720e725. 14. Cheung MMY, Saini B, Smith L. Using drawings to explore patients’ perceptions of their illness: a scoping review. J Multidiscip Healthc. 2016;9:631e646. 15. Levac D, Colquhoun H, O’Brien KK. Scoping studies: advancing the methodology. Implement Sci. 2010;5(1):69. 16. Crooks V, Kingsbury P, Snyder J, Johnston R. What is known about the patient’s experience of medical tourism? A scoping review. BMC Health Serv Res. 2010;10:266. 17. Piper BJ, Desrosiers CE, Lipovsky JW, et al. Use and misuse of opioids in Maine: results from pharmacists, the prescription monitoring, and the diversion alert programs. J Stud Alcohol Drugs. 2016;77(4):556e565. 18. Gavaza P, Fleming M, Barner JC. Examination of psychosocial predictors of Virginia pharmacists’ intention to utilize a prescription drug monitoring program using the theory of planned behavior. Res Social Adm Pharm. 2014;10(2):448e458. 19. Fleming ML, Barner JC, Brown CM, Shepherd MD, Strassels S, Novak S. Using the theory of planned behavior to examine pharmacists' intention to utilize a prescription drug monitoring program database. Res Social Adm Pharm. 2014;10(2):285e296. 20. Fass JA, Hardigan PC. Attitudes of Florida pharmacists toward implementing a state prescription drug monitoring program for controlled substances. J Manag Care Pharm. 2011;17(6):430e438. 21. Fleming ML, Phan Y, Ferries EA, Hatfield MD. Educating pharmacists on a prescription drug monitoring program. J Pharm Pract. 2016;29(6):543e548. 22. Ulbrich TR, Clark Dula CA, Green CG, Porter K, Bennett MS. Factors influencing community pharmacists' enrollment in a state prescription monitoring program. J Am Pharm Assoc. 2010;50:588e594.

23. Wixson SE, Blumenschein K, Goodin AJ, Talbert J, Freeman PR. Prescription drug monitoring program utilization in Kentucky community pharmacies. Pharm Pract. 2015;13(2):540. 24. Fendrich M, Hooijer K, Bryan J. Pharmacist PDMP use in a midwestern state implications for prevention and treatment. In: Fendrich M, Hooijer K, Bryan J, editors. American Public Health Association Conference. November 1, 2015; Chicago, IL. 25. Fleming ML, Chandwani H, Barner JC, Weber SN, Okoro TT. Prescribers and pharmacists requests for prescription monitoring program (PMP) data: Does PMP structure matter? J Pain Palliat Care Pharmacother. 2013;27(2):136e142. 26. Deyo RA, Irvine JM, Hallvik SE, Hildebran C, Beran T, Millet LM, et al. Leading a horse to water: Facilitating registration and use of a prescription drug monitoring program. Clin J Pain. 2014;31(9):782e787. 27. Norwood CW, Wright ER. Promoting consistent use of prescription drug monitoring programs (PDMP) in outpatient pharmacies: Removing administrative barriers and increasing awareness of Rx drug abuse. Res Social Adm Pharm. 2015;12(3):509e514. 28. Norwood CW, Wright ER. Integration of prescription drug monitoring programs (PDMP) in pharmacy practice: improving clinical decisionmaking and supporting a pharmacist's professional judgment. Res Social AdmPharm. 2016;12(2):257e266. 29. Green TC, Mann MR, Bowman SE, Zaller N, Soto X, Gadea J, et al. How does use of a prescription monitoring program change pharmacy practice? J Am Pharm Assoc. 2013;53(3):273e281. 30. Fleming ML, Barner JC, Brown CM, Shepherd MD, Strassels SA, Novak S. Pharmacists’ training, perceived roles, and actions associated with dispensing controlled substance prescriptions. J Am Pharm Assoc. 2014;54(3):241e250. 31. Rittenhouse R, Wei F, Robertson D, Ryan K. Utilization of the Arkansas prescription monitoring program to combat prescription drug abuse. Prev Med Rep. 2015;2:524e528. 32. Ajzen I. From intentions to actions: a theory of planned behavior. In: Kuhl J, Beckman J, eds. Action Control: From Cognition to Behavior. Heidelberg, Germany: Springer; 1985:11e39. 33. Brushwood DB. From confrontation to collaboration: collegial accountability and the expanding role of pharmacists in the management of chronic pain. J Law Med Ethics. 2001;29(1):69e93. 34. Borgsteede SD, Rhodius CA, De Smet PA, Pasman HR, OnwuteakaPhilipsen BD, Rurup ML. The use of opioids at the end of life: knowledge level of pharmacists and cooperation with physicians. Eur J Clin Pharmacol. 2011;67(1):79e89. 35. Hagemeier NE, Murawski MM, Lopez NC, Alamian A, Pack RP. Theoretical exploration of Tennessee community pharmacists’ perceptions regarding opioid pain reliever abuse communication. Res Social AdmPharm. 2014;10(3):562e575. 36. Deyo RA, Irvine JM, Millet LM, Beran T, O’Kane N, Wright DA, et al. Measures such as interstate cooperation would improve the efficacy of programs to track controlled drug prescriptions. Health Aff (Millwood). 2013;32(3):603e613. 37. Manasco AT, Griggs C, Leeds R, et al. Characteristics of state prescription drug monitoring programs: a state-by-state survey. Pharmacoepidemiol Drug Saf. 2016;25(7):847e851. 38. Ashburn MA. The evolution of prescription drug monitoring programs. Pharmacoepidemiol Drug Saf. 2016;25(7):852e853. 39. Rutkow L, Smith KC, Lai AY, Vernick JS, Davis CS, Alexander GC. Prescription drug monitoring program design and function: a qualitative analysis. Drug Alcohol Depend. 2017;180:395e400. 40. Doyle S, Leichtling G, Hildebran C, Reilly C. Research to support optimization of prescription drug monitoring programs. Pharmacoepidemiol Drug Saf. 2017;26(11):1425e1427. 41. Leichtling G, Hildebran C, Novak K, et al. Enhancing prescription drug monitoring program profiles: a qualitative study. In Preparation. 42. Bratberg JP. Opioids, naloxone, and beyond: the intersection of medication safety, public health, and pharmacy. J Am Pharm Assoc. 2017;57(2):S5eS7. 43. Zimmerman KM, Salgado TM, Goode J-VR, Sisson EM, Dixon DL. A prescription for prescribing: ensuring continued pharmacist preparedness. Ann Pharmacother. 2018;52(7):697e699. Kirbee Johnston, MPH, Faculty Research Assistant, Oregon State UniversityeOregon Health and Science University College of Pharmacy, Portland, OR Lindsey Alley, MS, Senior Research Associate, HealthInsight Oregon, Portland, OR Kevin Novak, MS, Research assistant, HealthInsight Oregon, Portland, OR Sarah Haverly, MS, Doctoral student, Portland State University, Portland, OR Adriane Irwin, MS, PharmD, BCACP, Assistant Professor (Clinical), Oregon State UniversityeOregon Health and Science University College of Pharmacy, Corvallis, OR Daniel Hartung, PharmD, MPH, Associate Professor, Oregon State UniversityeOregon Health and Science University College of Pharmacy, Portland, OR

9