What was the research about?
About 60 million people in the United States live with chronic pain that lasts for months or years. Doctors sometimes treat chronic pain with medicines called opioids. But patients who use opioids for a long time are at increased risk for serious health problems, such as addiction, overdose, car crashes, and injuries. The research isn’t clear if using opioids for a long time helps patients control their pain.
A network of clinics in Washington State started a program to reduce opioid doses prescribed for patients with chronic pain. The program had two phases:
- Dose-lowering. Doctors prescribed lower doses of opioids for many patients with chronic pain using opioids long-term. This phase started in 2007.
- Monitoring. Starting in 2010, clinics increased monitoring of patients taking opioids long-term and created care plans for them. Clinic pharmacies also changed the way patients refilled opioid prescriptions.
In this study, the research team wanted to learn if the program reduced the risks that go along with long-term opioid use. The study compared patients who received care at clinics with and without the program.
What were the results?
Patients who received care at clinics with and without the program had similar
- Rates of opioid overdoses
- Ratings of pain severity
- Rates of injuries and car crashes
During the dose-lowering phase, the rate of opioid overdoses decreased at clinics with the program, but not more than it did for clinics without the program.
Who was in the study?
The study looked at medical and public records for 31,142 adult patients who used opioids long-term. The research team also interviewed 1,588 of these patients.
What did the research team do?
The research team reviewed information on overdoses from patient records for three time periods. The first was before the dose-lowering phase started. The second was for the dose-lowering phase. The third was for the monitoring phase. The team also looked at numbers of car crashes and injuries. Four years after the program started, the team interviewed patients about their experience with pain. They compared patients from clinics with and without the program.
A group of patients and patient advocates helped design the study and review the results.
What were the limits of the study?
Policies about prescribing opioids changed in Washington State during the study period. The new policies may have changed how doctors’ prescribed opioids in both sets of clinics. Fewer patients received care at clinics without the program than at clinics with the program. As a result, it was hard to compare the groups of clinics.
Future research could study more people to better understand if reducing opioid doses helps lower the number of overdoses.
How can people use the results?
Clinics can use the results of this study when thinking about how to reduce opioid doses for patients with chronic pain.
To compare opioid doses and outcomes for patients with chronic pain receiving long-term opioid therapy at clinics implementing a program to lower opioid doses and increase monitoring versus patients in clinics that did not implement the program
|Design||Observational: cohort study|
|Population||EMRs of 31,142 adult patients receiving long-term opioid therapy for chronic pain management|
Primary: number of opioid overdoses
Secondary: patient-reported pain severity, number of injuries (fractures, concussions, and head injuries), number of motor vehicle crashes
|Timeframe||Up to 8-year follow-up for primary outcome|
This retrospective cohort study examined the effectiveness of a program that lowered opioid doses for patients receiving long-term opioid therapy for chronic pain. Primary care clinics within an integrated delivery system in Washington State implemented a program with a dose-reduction phase followed by an increased monitoring phase based on patient risk level. During the dose-reduction phase, which began in 2007, clinics with the program discouraged clinicians from escalating doses and using higher opioid doses. During the increased monitoring phase, which began in 2010, clinics with the program changed prescribing processes for opioids and increased clinician monitoring and development of patient-care plans. Affiliated clinics that were not part of the integrated delivery system did not implement this program and continued to offer usual care.
The research team reviewed electronic medical records (EMRs), medical claims data, and public records for 31,142 patients receiving long-term opioid therapy. The team retrospectively compared outcomes for patients from the clinics with and without the program during a baseline period prior to program implementation, the dose-reduction period, and the period of increased monitoring. Four years after program implementation began, the team surveyed a subset of 1,588 patients from the larger study to examine their experiences with pain.
A patient advisory committee that included patients, members of the American Chronic Pain Association, and advocates for reducing opioid overprescribing gave input on the study design and data analysis.
- Opioid overdoses. When comparing patients from clinics with and without the program, the research team did not find significantly different rates of opioid overdoses during either the dose-reduction or monitoring periods. However, opioid overdose rates decreased significantly each year during the dose-reduction period for patients from clinics with the program (p=0.04) but not for patients from clinics without the program.
- Pain severity. Four years after implementation, patients’ pain severity ratings were not significantly different between patients from clinics with and without the program.
- Injuries and collisions. Rates of injuries and motor vehicle crashes were similar during both the dose-reduction and monitoring periods for patients from clinics with and without the program.
Washington State introduced policies to reduce opioid doses during the time these programs took place. These statewide changes may have affected how doctors prescribed opioids in clinics with and without the program. The patient population in clinics without the program (n=8,469) was smaller than in clinics with the program (n=22,673), limiting the research team’s ability to detect significant differences between them.
Conclusions and Relevance
It is not clear whether the dose-reduction intervention did or did not lower overdose rates, because although the difference in overdose rates between clinics with and without the program did not differ, overdose rates decreased significantly during the dose-reduction phase for clinics with the program. Implementing the program did not change patient ratings of pain severity or reduce number of injuries or motor vehicle crashes.
Future Research Needs
Future research in larger populations could clarify whether reducing opioid doses lowers opioid overdose risk.
Final Research Report
View this project's final research report.
Results of This Project
Related Journal Citations
Stories and Videos
Peer review of PCORI-funded research helps make sure the report presents complete, balanced, and useful information about the research. It also assesses how the project addressed PCORI’s Methodology Standards. During peer review, experts read a draft report of the research and provide comments about the report. These experts may include a scientist focused on the research topic, a specialist in research methods, a patient or caregiver, and a healthcare professional. These reviewers cannot have conflicts of interest with the study.
The peer reviewers point out where the draft report may need revision. For example, they may suggest ways to improve descriptions of the conduct of the study or to clarify the connection between results and conclusions. Sometimes, awardees revise their draft reports twice or more to address all of the reviewers’ comments.
Peer reviewers commented, and the researchers made changes or provided responses. Those comments and responses included the following:
- Reviewers asked for more information about missing data, and specifically about whether missing data items might differ between the group practice and contracted-care clinics. The researchers responded that since the study data set contained the opioid prescription history of 96 percent of the patients in both groups, concerns about the effects of differences between groups would be limited.
- Reviewers questioned the generalizability of the study results since the researchers conducted the study within one health system, whereas a new state guideline would affect statewide opioid treatment decisions. The researchers acknowledged this concern but noted that the new state guideline was similar to the recommendations from the Centers for Disease Control and Prevention, which would make the results more relevant nationally than they would be without the guideline.
- Reviewers expressed concern about the choice of analysis for studying differences in adverse events based on dose reductions in the intervention and control groups. The analyses focused on the within-time segment slope—dose reduction, risk stratification and monitoring--without measuring the slope during the baseline period. The researchers explained that the dose-reduction slopes were very different in each time segment, so looking at the data across time segments might result in erroneous conclusions.
- Reviewers expressed concern about the timing of the study baseline period, which extended into the year after state opioid dosing guidelines changed. The researchers explained that they wanted to allow time for any initial, transient dose reductions related directly to the new guidelines to occur before starting their measurement. This also allowed time for the dose-reduction initiative to gain momentum.
- Reviewers expressed concern about how the researchers formed the study cohort, since individual patients could be eligible for inclusion during some quarters and ineligible for others, depending on their opioid prescriptions. Reviewers preferred a more stable sample that included the same patients throughout, marking them clearly as long-time opioid users. The researchers disagreed, stating that including only patients who were chronic users throughout would result in a much reduced and less-generalizable population, while keeping patients in the sample even if they no longer qualified, as chronic users would affect the validity of results. The open cohort employed here, meant that the researchers captured data for any patients in a particular quarter who met study criteria and their outcomes could be tracked for that period of time.
Conflict of Interest Disclosures
Study Registration Information
^Kaiser Permanente acquired Group Health Cooperative in February, 2017.
- Has Results