Article Figures & Data
Tables
Urban, Underserved Clinics Suburban Clinics Ranking Within Overall Prescriptionsa Medication Name No. (Percentage of Total Prescriptions) Ranking Within Overall Prescriptionsa Medication Name No. (Percentage of Total Prescriptions) 6 Oxycodone 3,365 (2.5) 47 Codeine-guaifenesin 2,271 (0.6) 8 Acetaminophen-oxycodone 2,873 (2.1) 56 Hydrocodone-acetaminophen 2,039 (0.5) 13 Acetaminophen-hydrocodone 2,074 (1.6) 60 Hydrocodone-chlorpheniramine 1,873 (0.5) 15 Tramadol 1,855 (1.4) 61 Oxycodone-acetaminophen 1,855 (0.5) 61 Fentanyl 554 (0.4) 68 Tramadol 1,680 (0.4) 85 Morphine 418 (0.3) 102 Oxycodone 865 (0.2) 88 Codeine-guaifenesin 398 (0.3) 134 Acetaminophen-codeine 621 (0.2) 95 Methadone 368 (0.3) 196 Morphine 295 (0.1) 117 Chlorpheniramine-hydrocodone 291 (0.2) 232 Promethazine-codeine 221 (0.1) 131 Acetaminophen-codeine 246 (0.2) 236 Hydromorphone 210 (0.1) ↵a This ranking compares the top 10 prescribed opioids with all prescriptions (including nonopioid prescriptions) at 21 primary care clinics. Note that opioids can be prescribed only at monthly intervals without refills, whereas other medications may have refills that can last for up to 1 year.
- Table 2
Demographic Characteristics and Risk Profiles for Opioid-Related Harms of Patients Receiving Chronic Opioids
Patients on Chronic Opioids, n = 902 (%)a Patients Not on Chronic Opioids, n = 84,027 (%)a P Value Demographic characteristics Sex, female 591 (65.5%) 50,575 (60.2%) <.01 Race Asian/Asian American 6 (0.7%) 6,034 (7.2%) <.01 Black/African American 389 (43.1%) 12,377 (14.7%) White 435 (48.2%) 48,200 (57.4%) Other 20 (2.2%) 4,509 (5.4%) Declined 52 (5.8%) 12,907 (15.4%) Ethnicity, Hispanic 12 (1.4%) 3,576 (5.4%) <.01 Medical comorbidities Sleep apnea 194 (21.5%) 7001 (8.3%) <.01 Depression or anxiety 537 (59.5%) 23,693 (28.2%) <.01 Substance use disorder 118 (13.1%) 1,802 (2.1%) <.01 Hepatic insufficiency 153 (17.0%) 4,020 (4.8%) <.01 Renal insufficiency 174 (19.3%) 2,439 (2.9%) <.01 Concurrent benzodiazepine prescription 216 (24.0%) 714 (8.3%)b <.01 ↵a The denominator for all percentages is the n listed in the column heading unless otherwise indicated.
↵b Because concurrent benzodiazepine prescription is relevant only for patients with an opioid prescription, the denominator here is 8,560 to account for patients who have received an opioid prescription in 2016.
- Table 3
Variation in Daily Morphine Milligram Equivalent (MME) for Patients on Chronic Opioids by Patient, Clinician, and Clinic Characteristics
Characteristics N Median Daily MME (IQR) Kruskal-Wallis P Value Patient level Sex Female 591 30.0 (15.0-51.4) .06 Male 311 30.0 (15.9-60.0) Race Asian/Asian American 6 13.7 (5.8-36.3) <.01 Black/African 389 30.0 (20.0-52.5) American White 435 30.0 (13.7-60.0) Other 20 24.2 (11.3-59.6) Declined 52 13.1 (5.7-30.7) Ethnicity Hispanic 12 20.6 (5.8-49.4) .08 Not Hispanic 820 30.0 (15.7-57.2) Sleep apnea Yes 194 30.0 (20.0-57.2) .08 No 708 30.0 (15.0-55.5) Depression or anxiety Yes 537 31.5 (16.7-60.0) <.01 No 365 25.6 (15.0-48.3) Substance use disorder Yes 118 41.4 (22.5-82.5) <.01 No 784 30.0 (15.0-52.5) Hepatic insufficiency Yes 153 38.9 (18.2-60.0) .01 No 749 30.0 (15.0-53.7) Renal insufficiency Yes 174 30.0 (20.0-60.0) .03 No 728 30.0 (15.0-54.0) Concurrent benzodiazepine prescription Yes 216 31.0 (15.0-69.6) .09 No 686 30.0 (15.0-51.1) Clinician level Degree type MD 157 43.3 (19.1-67.7) <.01 DO 10 38.6 (26.3-57.9) NP 21 18.4 (7.1-57.0) PA 7 5.1 (3.0-14.6) Other 1 20.0 (20.0-20.0) Age <50 137 43.4 (18.2-61.7) .24 ≥50 43 35.1 (14.0-56.5) Sex Female 113 33.6 (11.2-53.2) <.01 Male 83 46.8 (25.1-69.7) Specialty Family medicine 103 18.2 (7.5-42.0) <.01 Internal medicine 93 51.0 (40.1-74.2) Trainee or attending status Attending 131 25.4 (9.2-56.5) <.01 Resident 65 52.5 (40.2-71.0) Clinic level Type Urban, underserved 8 61.6 (54.4-65.4) .05 Suburban 10 26.1 (15.0-57.6) Number of clinicians <5 clinicians 5 25.1 (13.7-51.0) .26 5+ clinicians 13 57.7 (28.3-64.1) Training site? Residents 2 68.3 (64.1-72.5) .09 No residents 16 40.0 (22.1-61.6) Themes Findings Example Quotes Inheriting patients on chronic opioids Clinicians feel frustrated after inheriting patients with chronic pain on opioids. “Her pain doctor got fired or, you know, loss of license (I can’t remember what happened) so she came to me very angry, very difficult conversation wanting to be on the same huge doses.” Clinicians feel pressured to manage inappropriate amounts or types of chronic opioids. “I get a call from the case manager who says would you be willing to consider the following: Would you be willing to consider prescribing her methadone as a pain management medication and not as methadone maintenance for her substance abuse disorder…. Looking back now it’s one of these …what the heck was I thinking kind of thing…. I sort of felt like I was getting stiff-armed into things.” Primary care often becomes the default chronic pain medication manager. “She is seeing a psychiatrist, a pain specialist, an orthopedist, and a rheumatologist. She’s got all of these people involved in her care but, for some reason, I’m the person who stuck with her pain med management and nobody is super-eager to touch that.” Co-occurring health problems Extent or complexity of medical comorbidities often takes priority. “He’s kind of a mess…. Bad asthma. Bad COPD. Bad heart. He’s been in the hospital more than he’s been out and was diagnosed with leukemia about six months ago. I told him, you know, your pain pills are the least of my concerns; whatever it takes to get you through the day.” Patients with chronic pain often have complex social situations. “She’s someone I’ve been trying to wean but her social situation is just a disaster. She’s leaving an abusive relationship. I think she’s actually homeless right now.” Contraindication of nonopioid pain medications limits pain management strategies. “I will say that the older ones …have crappy kidneys and contraindications to a lot of other things that you could actually put them on to help with their pain.” Access to adjunctive pain management strategies is limited. “She’s stable and is a single mom working and has limited access to adjunctive therapies and physical therapies. I’m a little stuck on her.” Benefits of opioids for chronic pain management Chronic opioids are necessary to sustain functional capacity and quality of life. “I feel like a change is not indicated at this time because she needs the medication in order to do her job and go to work and help her family, and it is working for her. She is overall low-risk for abuse. I don’t feel compelled to make a change for her.” Chronic opioids help manage chronic diseases. “She’s got chronic pancreatitis and without this dose she can’t eat and she gets malnourished and gets really sick.” Challenges with weaning Clinicians lack time to manage chronic opioids. “It might be different if I were a chronic pain management doctor and I was seeing 20 patients a day every day and just doing this.” Lack of control over other sources of opioids undermines weaning attempts. “The amount of narcotics she gets in the hospital blows us out of the water. The problem is she is on around-the-clock high-dose IV-pain medicine when she goes to the hospital.” Clinicians have a hard time justifying weaning stable patients receiving long-term opioid pain management. “It’s very hard with patients who have been on them [a long time] to get them off…. Five to seven years ago, [when many of my patients started on opioid], this wasn’t even part of the cultural discussion. It was like patients just came in and demanded these medications, they could be downright hostile with you. Now, I could tell them we’re going to wean down, but it would be a really ugly fight.”
Additional Files
The Article in Brief
Chronic Opioid Prescribing in Primary Care: Factors and Perspectives
Sebastian Tong , and colleagues
Background Primary care clinicians write almost half of all opioid prescriptions in the United States but little is known about the characteristics of these clinicians and their patients who receive opioids. This study describes patient and clinician characteristics and clinicians' perspectives of chronic opioid prescribing in primary care
What This Study Found Chronic opioid prescribing in primary care varies significantly by patient and clinician characteristics. This analysis of 2016 electronic health record data included 21 primary care practices, 271 clinicians, and 84,929 patients. Eleven percent of patients seen received an opioid prescription, while 1% received chronic opioid prescriptions. Oxycodone-acetaminophen was the most commonly prescribed opioid, followed by oxycodone. In urban underserved clinics, 10% of prescriptions written were for opioids, compared to 3% of prescriptions in suburban clinics. Being female, being of black race, and having risks for opioid-related harms, such as mental health diagnoses, substance use disorder and concurrent benzodiazepine use, were associated with being prescribed chronic opioids. Patients with higher comorbidities were more likely to receive chronic opioid prescriptions and at higher doses. In interviews, clinicians described the use of opioids to manage chronic pain as appropriate for patients with extensive medical comorbidities or patients for whom non-opioid pain medications were contraindicated. However, most were reluctant to begin patients on opioids for chronic pain. Many felt frustrated by lack of time to appropriately manage patients� chronic pain and lack of control over patients' access to other sources of opioids.
Implications
- The authors call for research to explore trends in opioid prescribing, compare the differences in opioid prescribing in various settings, and test interventions to help primary care clinicians overcome barriers in weaning patients with high risks of opioid-related harms.