Clinical Styles and Practice Policies: Influence on Communication with Patients Regarding Worrisome Prescription Drug Monitoring Program Data

Abstract

Objectives. Clinician communication with patients regarding worrisome findings in Prescription Drug Monitoring Programs (PDMPs) may influence patient responses and subsequent care. The authors studied the range of approaches clinicians report when communicating with patients in this situation and how practice policies and procedures may influence this communication.

Design. Qualitative interviews of clinician PDMP users.

Setting: Oregon registrants in the state’s PDMP.

Subjects. Thirty-three clinicians practicing in pain management, emergency medicine, primary care, psychiatry, dentistry, and surgery.

Methods. The authors conducted semi-structured interviews via telephone with clinicians who routinely used the PDMP. A multidisciplinary team used a grounded theory approach to identify ways clinicians reported using information from the PDMP when communicating with patients, and policies that influenced that communication.

Results. Clinicians reported using a range of approaches for communicating about PDMP results, from openly sharing, to questioning patients without disclosing access to the PDMP, to avoiding the conversation. Clinicians also reported practice policies and procedures that influenced communication with their patients about prescribing and ongoing care, including policies that normalized use of the PDMP with all patients and those that facilitated difficult conversations by providing a rationale not to prescribe in certain circumstances.

Conclusion. Clinicians’ self-reported approaches to sharing PDMP findings and communicating prescribing decisions with patients vary and may be facilitated by appropriate practice policies. Such communication may have implications for patient engagement and alliance building. More research is needed to identify best practices and potential guidelines for effectively communicating about PDMP findings, as this may enhance health outcomes.

Introduction

Prescription drug monitoring programs (PDMPs) have become operational in most states over the past decade, intended to aid clinicians in identifying patients with possible prescription drug misuse, abuse, addiction, harmful use, or diversion [1]. These tools help identify worrisome PDMP profiles, which have been described as those that indicate patients have prescriptions from multiple providers, risky drug combinations (e.g., opioid and benzodiazepine prescription), or high opioid doses [2]. PDMPs may also increase clinician confidence in prescribing for patients who are most likely to benefit from pain medication.

Conversations about chronic pain management and use of prescription opioids are often difficult and frustrating for both clinicians and patients [3–5], and little is known about how the PDMP influences these conversations. In our previous focus groups with providers in nine states [6], participants reported discussing worrisome PDMP profiles with patients in a variety of ways; we found similar results in our interviews with Oregon clinicians. Focus groups and surveys of clinician PDMP users identified experiences of both positive and negative patient responses to discussion of worrisome PDMP profiles, with some reporting patient anger or denial [7]. Still, little is known about how clinicians integrate discussion of worrisome PDMP profiles into their visits with patients, or whether the approach clinicians use influences patient responses. This report contributes to the body of knowledge about clinician communication of PDMPs.

Policies, guidelines, and continuing education related to opioid prescribing at the federal, state, and local levels refer to use of the PDMP [8,9] and are intended to foster use of PDMP data in making prescribing decisions. A recent study of emergency physicians [10] found that physicians primarily used prescribing guidelines as a communication tool rather than a decision-making tool. It remains to be seen how such general prescribing policies and guidelines intersect with individual patient PDMP profiles to influence communication about prescribing. To examine these issues in greater depth, we conducted semi-structured interviews with users of Oregon’s PDMP. Specifically, we sought to understand 1) how clinicians discuss worrisome PDMP profiles with patients, 2) whether policies related to the PDMP or opioid prescribing influence clinician communication about the PDMP, and 3) clinician perceptions of how patients respond to discussion of worrisome PDMP profiles.

Methods

The Oregon Public Health Division and Oregon Health & Science University Institutional Review Boards approved this study. In addition, we obtained a Certificate of Confidentiality from the National Institutes of Health in order to protect participant confidentiality.

Sample

Registered users of the PDMP who completed an Oregon clinician survey in 2013 were eligible for participation in this study (n = 619). Of the 619 survey respondents, 212 provided follow-up contact information. In addition to practice specialty and reported willingness to be contacted, we used three attributes, based on clinician survey responses, to purposively select a maximum variation sample of clinicians for follow-up telephone interviews. These attributes were: 1) practice setting (i.e., primary care, emergency medicine, pain, psychiatry, dentistry, etc.), 2) experiences with patients related to PDMP use (e.g., those reporting that patients rarely respond in denial or anger, reported patients walking out of a visit, reported discharging patients from their practice because of PDMP information), and 3) open-ended comments suggesting a unique clinic process or conversation strategy. We purposively selected 60 respondents and 33 consented and completed an interview. Those interviewed represented wide variation on the attributes of interest, as shown in Table 1. Interviewees and non-respondents did not differ substantially in professional or demographic characteristics.

Table 1.

Characteristics of the 33 interview participants

	Specialty
	Primary Care	Emergency Medicine	Dental	Surgical	Pain Medicine	Addictions/Psychiatry	Total
Gender
Female	6	2	3		1	2	14
Male	10	4	2	2	1		19
Credential
MD/DO	9	3			2	2	16
NP	4	2					6
PA	3	1		2			6
Dentist			5				5
Total	16	6	5	2	2	2	33

Recruitment

A study team member contacted clinicians by phone. Of the 60 contacted, 33 provided verbal consent over the phone and scheduled the interview. After an interview was completed, participants were mailed a $150 check as compensation for their time.

Data Collection and Management

Interviews followed a semi-structured guide with questions focusing on use of the PDMP, related clinical processes and workflows, policies, and strategies for discussing PDMP results with patients. For the latter, clinicians were asked to describe examples of actual discussions with patients. Prior focus group research with clinicians suggested that PDMP use might vary depending on whether the patient was a new or existing patient [6]. We designed the interview guide to explore this distinction and how it might affect clinical decisions, actions, and patient interactions. The interview guide is available in the appendix.

Data collection and analysis was iterative, with some interviews being conducted and then analyzed by the larger team before collecting additional interviews. This allowed our team to modify the interview guide slightly to elicit specific examples or stories of clinicians’ encounters with patients when the PDMP was used (see Attachment A) and to monitor when saturation was reached, the point at which no new themes were emerging from the data collection process. Saturation was reached with the completion of 33 interviews.

Researchers with experience in qualitative interviewing methods conducted the telephone interviews. Interviews were conducted in pairs, with one person leading the interview and a second person taking notes and asking follow-up questions at the end. Participants were assigned a five digit participant number and participant names and identifiers were not used during the interview. Interviews were digitally recorded and professionally transcribed by uploading to a secure site. Transcriptions were reviewed for accuracy, de-identified and then uploaded into ATLAS.ti™ software (version 7.1.3) for qualitative analysis.

Analysis

A multidisciplinary team composed of a primary care physician, clinical pharmacist, addiction therapist, communication scientist, and two qualitative research analysts reviewed transcripts. We used a grounded theory approach [11] and engaged in two immersion-crystallization cycles to analyze data [12]. In the first cycle, our team read five participant transcripts aloud as a group to gain insight into participants’ experiences and use of PDMP (immersion). Through this process we identified important segments of text and identified key emerging findings (crystallization) that we developed into a code list for use in analyzing subsequent interviews. Thereafter, three team members analyzed the remaining interviews, checking in regularly, while using and refining codes as findings and new insights emerged. Preliminary findings were continually discussed with a fourth research team member, and periodically reviewed with the multidisciplinary team to further clarify findings.

Results

The following describe clinician’s self-reported approaches to discussing worrisome PDMP profiles, whether policies influence communication and perspective of patient responses.

How Clinicians Discuss Worrisome PDMP Profiles with Patients

Clinicians reported discussing worrisome PDMP results with patients in a variety of ways. One strategy discussed was to describe the worrisome profile and express safety concerns:

For me, it’s a chance to broach the topic in an objective, non-judgmental way, to say, “You’re 20 years old and I see you’ve gotten 160 Vicodin over the last month. This is dangerous. This is not good medicine. I’m concerned about this.” (Participant 1, emergency room clinician)

Other clinicians reported asking patients to describe their prescription history without first revealing that the clinician knew of a worrisome PDMP profile. In this approach, the PDMP was used as a tool to detect patient dishonesty:

It’s a cat and mouse thing. So I keep it [the PDMP] secret as much as possible, because it’s better used if it’s kept quiet. I can catch the patient unaware. I would never say “I’m monitoring you and you’re not going to be able to get away with it.” It’s much better for me to have information and I can discover things that are happening. You have to be a bit of a detective. If people aren’t going to be truthful you have to use all the resources you can. (Participant 2, primary care clinician)

Some clinicians also described avoiding direct discussion of the worrisome PDMP profile and instead invoked clinic policy against prescribing:

I never confront them with the evidence from the PDMP. I write it up in the chart, so the chart indicates … 18 prescriptions for controlled substances, from six different providers over the last year, or whatever. I’ll flag his chart as drug seeking and that will be his number one diagnosis. And I’ll write in the chart, “Do not prescribe controlled substances,” in large capital letters. But I won’t confront him or her with it. I’ll just say, “You know, clinic policy, can’t prescribe, they’re not safe for you, but we’ll help you.” (Participant 3, primary care clinician)

Whether Policies Related to PDMP or Prescribing Influence Communication

Clinicians reported discussing policies with patients in two situations related to PDMP. First, they described invoking policies as a way of explaining their use of the PDMP prior to prescribing. Second, they invoked policies as a reason not to prescribe a controlled medication following a worrisome PDMP profile.

A primary care clinician reported explaining the use of PDMP in the context of clinic policies that included a number of required practices such as use of a medication contract:

I tell them that as part of our clinic policy, I need to log in and see if they're getting prescribed these medications from another prescriber. It helps me to tell them about the policy, the medication contract that we have them sign if they do get them [controlled prescriptions] long term from us. (Participant 4)

An emergency room clinician reported communicating to patients with a worrisome PDMP profile that the decision not to prescribe was based on policy that prohibits prescribing opioids for chronic pain or lost medications:

I just explain that I have information that’s verifiable, that would contraindicate what they’ve said [about their prescription history]. I explain that we as an emergency room, and a community working together, have established a policy and guideline that we don’t believe it’s helpful to refill lost or stolen recurrent prescriptions for opiate pain medicines, and it’s neither safe nor effective nor in the patient’s best interest to prescribe opiate pain medicines for recurrent long-lasting pain problems. (Participant 5)

Clinician Perceptions of Patient Responses

Clinicians reported a range of patient responses to discussion of worrisome PDMP profiles, including surprise, anger, embarrassment, and occasionally acknowledgement of opioid use problems and acceptance of the clinician’s prescribing decision. As noted by a pain clinician:

There are as many different responses as you can imagine. “How do you expect me to get my pain medicine if you won’t give them to me?” to “I have bad teeth” or “I was in Utah at my daughter’s wedding and I left my pills at home.” (Participant 6)

An emergency room clinician described the unpredictability of patient responses:

Some people accept the information rather quickly, without being upset or getting agitated. Some people push back and start negotiating. It is somewhat unpredictable and a varied set of responses. Certainly some people rapidly escalate, and sometimes you have to call security so they can be shown off campus. But that’s a minority of cases. (Participant 5)

Even when clinicians reported a non-confrontational, supportive approach to discussing worrisome PDMP profiles, some noted that patients may have an unfavorable response to their communication.

People respond the same whether I do it early in the relationship or later, whether I do it with evidence or not with evidence. It doesn’t really matter what I have, that anger response seems to happen. They’re people who are using kind of inappropriately, it doesn’t seem to matter how I approach them, they’re just mad. (Participant 7, pain clinician)

On the other hand, one clinician described how the conversation can lead to a positive patient response once alternatives to prescribing were discussed.

I’ll tell them that there are a lot of other great treatments for pain, so we’ll start exploring these other treatments and start implementing them. And very often, usually they’re pretty happy with that. (Participant 8, primary care clinician)

Discussion

Clinician participants expanded on, in important ways, the communication approaches identified in prior focus group studies on this topic. Some clinicians reported that the PDMP helps them to have an objective “non-judgmental” conversation with patients, while others reported avoiding discussions of worrisome PDMP profiles altogether. Still others described using the PDMP as a way to detect patient dishonesty. We speculate that the latter approach may foster defensive reactions or weaken the therapeutic alliance, a factor that appears to influence treatment outcomes [13].

Research shows that clinicians report substantial frustration and a heavy emotional toll in caring for patients with chronic pain, particularly in relation to opioid therapy [5,14]. This toll may influence clinicians’ responses to worrisome PDMP profiles. Nonetheless, clinicians in our study report constructive approaches to discussing these reports—for example, sharing the findings and expressing safety concerns. Even if a clinician has no intention of prescribing or continuing a prescription, such a conversation may help identify specific problems and courses of action. Patient-clinician communication at these delicate junctures may also affect patient retention in care. Patient-physician relationship can affect outcomes [13,15], highlighting a need for strategies to improve patient engagement and retention and to foster open communication regarding prescribed controlled drugs.

Clinicians described discussing clinic policies related to accessing the PDMP or to prescribing decisions following a worrisome PDMP report. Invoking policy is a conversational strategy for framing the clinician’s process and decision as standard operating procedure, thus placing responsibility with a higher authority (e.g., clinic or professional association) and bolstering support for the clinician’s actions. Patients may interpret such policies as normalizing in that the individual patient is not targeted or treated with bias. On the other hand, clinicians may avoid conflict by communicating that decisions were based on prescribing policies and may miss an opportunity to discuss the particular concerns (e.g., need for addiction therapy or mental health referral) raised by a patient’s PDMP profile. Recent literature suggests that a patient’s desire to be treated as an individual by his or her provider is important in pain management, as is acknowledgement of patients’ individual needs and risks [3].

Clinicians detailed a range of patient responses to discussing worrisome PDMP profiles, from accepting the clinician’s decisions to attempting to negotiate for pain medications to becoming angry. While studies suggest that patients are more receptive to clinicians’ decisions to deny or reduce opioids when they believe the clinician is genuinely concerned for their health [16], it is unclear to what extent patient response is influenced by the clinicians’ communication approach or which approach may be optimal in working with patients. Some of the clinicians reported negative patient responses no matter the approach. However, this self-perception and self-report may be influenced by a range of recall biases. Thus, naturalistic observation studies are needed.

A strength of this study is that all participants were regular users of the PDMP, and were purposively selected to maximize variation in specialty and use of the PDMP system. Thus, a broad range of perspectives and practices enrich the study results. This study also has important limitations. We relied on clinician self-report, which may not reflect actual or typical communication with patients. We selected regular users of the PDMP in order to reach clinicians with significant experience communicating with patients about PDMP. Clinicians with less regular use of the PDMP may have different communication patterns. Other limitations include potential biases of the researchers. We mitigated this by involving a multidisciplinary team in data analysis.

Although PDMPs are valuable tools to better understand a patient’s prescription history, the way this information is discussed with patients may influence whether the patient engages in appropriate care. A worrisome PDMP profile likely indicates the need for medical attention—for example, referral to addiction treatment or mental health care, a change in prescription type or dosage, or education related to safe medication use. Guidance based on communications research, including direct observation, may facilitate diagnosis and treatment and more successful engagement of patients in their care.