Prevalence of Prescription Opioid Misuse/Abuse as Determined by International Classification of Diseases Codes: A Systematic Review

Sensitivity and specificity of International Classification of Diseases algorithms (ICD-9 and ICD-10) used to identify opioid-related overdose cases: A systematic review and an example of estimation using Bayesian latent class models in the absence of gold standards

OBJECTIVES

This study aimed to summarize validity estimates of International Classification of Diseases (ICD) codes in identifying opioid overdose (OOD) among patient data from emergency rooms, emergency medical services, inpatient, outpatient, administrative, medical claims, and mortality, and estimate the sensitivity and specificity of the algorithms in the absence of a perfect reference standard.

METHODS

We systematically reviewed studies published before December 8, 2023, and identified with Medline and Embase. Studies reporting sufficient details to recreate a 2 × 2 table comparing the ICD algorithms to a reference standard in diagnosing OOD-related events were included. We used Bayesian latent class models (BLCM) to estimate the posterior sensitivity and specificity distributions of five ICD-10 algorithms and of the imperfect coroner's report review (CRR) in detecting prescription opioid-related deaths (POD) using one included study.

RESULTS

Of a total of 1990 studies reviewed, three were included. The reported sensitivity estimates of ICD algorithms for OOD were low (range from 25.0% to 56.8%) for ICD-9 in diagnosing non-fatal OOD-related events and moderate (72% to 89%) for ICD-10 in diagnosing POD. The last included study used ICD-9 for non-fatal and fatal and ICD-10 for fatal OOD-related events and showed high sensitivity (i.e. above 97%). The specificity estimates of ICD algorithms were good to excellent in the three included studies. The misclassification-adjusted ICD-10 algorithm sensitivity estimates for POD from BLCM were consistently higher than reported sensitivity estimates that assumed CRR was perfect.

CONCLUSION

Evidence on the performance of ICD algorithms in detecting OOD events is scarce, and the absence of bias correction for imperfect tests leads to an underestimation of the sensitivity of ICD code estimates.

Identifying patients with opioid use disorder using International Classification of Diseases (ICD) codes: Challenges and opportunities

BACKGROUND AND AIMS

International Classification of Diseases (ICD) diagnosis codes are often used in research to identify patients with opioid use disorder (OUD), but their accuracy for this purpose is not fully evaluated. This study describes application of ICD-10 diagnosis codes for opioid use, dependence and abuse from an electronic health record (EHR) data extraction using data from the clinics' OUD patient registries and clinician/staff EHR entries.

DESIGN

Cross-sectional observational study.

SETTING

Four rural primary care clinics in Washington and Idaho, USA.

PARTICIPANTS

307 patients.

MEASUREMENTS

This study used three data sources from each clinic: (1) a limited dataset extracted from the EHR, (2) a clinic-based registry of patients with OUD and (3) the clinician/staff interface of the EHR (e.g. progress notes, problem list). Data source one included records with six commonly applied ICD-10 codes for opioid use, dependence and abuse: F11.10 (opioid abuse, uncomplicated), F11.20 (opioid dependence, uncomplicated), F11.21 (opioid dependence, in remission), F11.23 (opioid dependence with withdrawal), F11.90 (opioid use, unspecified, uncomplicated) and F11.99 (opioid use, unspecified with unspecified opioid-induced disorder). Care coordinators used data sources two and three to categorize each patient identified in data source one: (1) confirmed OUD diagnosis, (2) may have OUD but no confirmed OUD diagnosis, (3) chronic pain with no evidence of OUD and (4) no evidence for OUD or chronic pain.

FINDINGS

F11.10, F11.21 and F11.99 were applied most frequently to patients who had clinical diagnoses of OUD (64%, 89% and 79%, respectively). F11.20, F11.23 and F11.90 were applied to patients who had a diagnostic mix of OUD and chronic pain without OUD. The four clinics applied codes inconsistently.

CONCLUSIONS

Lack of uniform application of ICD diagnosis codes make it challenging to use diagnosis code data from EHR to identify a research population of persons with opioid use disorder.

Sensitivity of Medicaid Claims Data for Identifying Opioid Use Disorder in Patients Admitted to 6 New York City Public Hospitals

OBJECTIVES

Behavioral health diagnoses are frequently underreported in administrative health data. For a pragmatic trial of a hospital addiction consult program, we sought to determine the sensitivity of Medicaid claims data for identifying patients with opioid use disorder (OUD).

METHODS

A structured review of electronic health record (EHR) data was conducted to identify patients with OUD in 6 New York City public hospitals. Cases selected for review were adults admitted to medical/surgical inpatient units who received methadone or sublingual buprenorphine in the hospital. For cases with OUD based on EHR review, we searched for the hospitalization in Medicaid claims data and examined International Classification of Diseases, Tenth Revision discharge diagnosis codes to identify opioid diagnoses (OUD, opioid poisoning, or opioid-related adverse events). Sensitivity of Medicaid claims data for capturing OUD hospitalizations was calculated using EHR review findings as the reference standard measure.

RESULTS

Among 552 cases with OUD based on EHR review, 465 (84.2%) were found in the Medicaid claims data, of which 418 (89.9%) had an opioid discharge diagnosis. Opioid diagnoses were the primary diagnosis in 49 cases (11.7%), whereas in the remainder, they were secondary diagnoses.

CONCLUSION

In this sample of hospitalized patients receiving OUD medications, Medicaid claims seem to have good sensitivity for capturing opioid diagnoses. Although the sensitivity of claims data may vary, it can potentially be a valuable source of information about OUD patients.

An Evaluation of the Effect of the OxyContin Reformulation on Unintentional Fatal and Nonfatal Overdose

OBJECTIVES

OxyContin was reformulated with a polyethylene oxide matrix in August 2010 to reduce the potential for intravenous abuse and for abuse by insufflation. The objective of this study was to evaluate the impact of OxyContin's reformulation on overdose (OD) risk for individuals dispensed OxyContin in comparison to those dispensed other opioids under regular care.

MATERIALS AND METHODS

Three national insurance databases with National Death Index linkage identified OD in individuals with any dispensing of OxyContin or a primary comparator opioid (extended release morphine, transdermal fentanyl, or methadone) between July 2008 through September 2015. A difference-in-differences design was used to compare the pre-post reformulation changes in OD rates for OxyContin versus comparators.

RESULTS

A total of 297,836 individuals were dispensed OxyContin and 659,673 individuals were dispensed a primary comparator across the 3 databases. Overall, there was little or no difference in the temporal change in OD incidence in comparators versus OxyContin (Medicaid: adjusted ratio-of-rate-ratios (aRoRs) ranging from 0.90 to 1.05; MarketScan/HIRD: aRoR ranging from 1.10 to 1.22). However, restriction to person-time without concomitant opioid use revealed a modestly greater reduction in OD incidence over time during OxyContin use, as the aRoRs comparing the primary comparators to OxyContin ranged from 1.06 to 1.30 in Medicaid and from 1.64 to 1.85 in MarketScan/HIRD.

DISCUSSION

This study did not detect an overall effect of the OxyContin reformulation on OD in insured patients under regular medical care. There is a suggestion of a modestly reduced OxyContin-associated OD risk following the reformulation but only in commercially insured individuals receiving single-opioid regimens.

Automatically identifying opioid use disorder in non-cancer patients on chronic opioid therapy

Background: Using the International Classification of Diseases (ICD) codes alone to record opioid use disorder (OUD) may not completely document OUD in the electronic health record (EHR). We developed and evaluated natural language processing (NLP) approaches to identify OUD from the clinal note. We explored the concordance between ICD-coded and NLP-identified OUD.Methods: We studied EHRs from 13,654 (female: 8223; male: 5431) adult non-cancer patients who received chronic opioid therapy (COT) and had at least one clinical note between 2013 and 2018. Of eligible patients, we randomly selected 10,218 (75%) patients as the training set and the remaining 3436 patients (25%) as the test dataset for NLP approaches.Results: We generated 539 terms representing OUD mentions in clinical notes (e.g., "opioid use disorder," "opioid abuse," "opioid dependence," "opioid overdose") and 73 terms representing OUD medication treatments. By domain expert manual review for the test dataset, our NLP approach yielded high performance: 98.5% for precision, 100% for recall, and 99.2% for F-measure. The concordance of these NLP and ICD identified OUD was modest (Kappa = 0.63).Conclusions: Our NLP approach can accurately identify OUD patients from clinical notes. The combined use of ICD diagnostic code and NLP approach can improve OUD identification.

Opioid-related emergency department visits and hospitalizations among patients with chronic gastrointestinal symptoms and disorders dually enrolled in the Department of Veterans Affairs and Medicare Part D

PURPOSE

We examined the prevalence of, and factors associated with, serious opioid-related adverse drug events (ORADEs) that led to an emergency department (ED) visit or hospitalization among patients with chronic gastrointestinal (GI) symptoms and disorders dually enrolled in the Department of Veterans Affairs (VA) and Medicare Part D.

METHODS

In this retrospective cohort study, we used linked national patient-level data (April 1, 2011, to October 31, 2014) from the VA and Centers for Medicare and Medicaid Services to identify serious ORADEs among dually enrolled veterans with a chronic GI symptom or disorder. Outcome measures included serious ORADEs, defined as an ED visit attributed to an ORADE or a hospitalization where the principal or secondary reason for admission involved an opioid. We used multiple logistic regression models to determine factors independently associated with a serious ORADE.

RESULTS

We identified 3,430 veterans who had a chronic GI symptom or disorder; were dually enrolled in the VA and Medicare Part D; and had a serious ORADE that led to an ED visit, hospitalization, or both. The period prevalence of having a serious ORADE was 2.4% overall and 4.4% among veterans with chronic opioid use (≥90 consecutive days). Veterans with serious ORADEs were more likely to be less than 40 years old, male, white, and to have chronic abdominal pain, functional GI disorders, chronic pancreatitis, or Crohn's disease. They were also more likely to have used opioids chronically and at higher daily doses.

CONCLUSION

There may be a considerable burden of serious ORADEs among patients with chronic GI symptoms and disorders. Future quality improvement efforts should target this vulnerable population.

Identifying individuals with opioid use disorder: Validity of International Classification of Diseases diagnostic codes for opioid use, dependence and abuse

BACKGROUND

Policy evaluations and health system interventions often utilize International Classification of Diseases (ICD) codes of opioid use, dependence, and abuse to identify individuals with opioid use disorder (OUD) and assess receipt of evidence-based treatments. However, ICD codes may not map directly onto the Diagnostic and Statistical Manual of Mental Disorder (DSM-5) OUD criteria. This study investigates the positive predictive value of ICD codes in identifying patients with OUD.

METHODS

We conducted a clinical chart review on a national sample of 520 Veterans assigned ICD-9 or ICD-10 codes for opioid use, dependence, or abuse from 2012 to 2017. We extracted evidence of DSM-5 OUD criteria and opioid misuse from clinical documentation in the month preceding and three months following initial ICD code listing, and categorized patients into: 1) high likelihood of OUD, 2) limited aberrant opioid use, 3) prescribed opioid use without evidence of aberrant use, and 4) insufficient information. Positive predictive value was calculated as the percentage of individuals with these ICD codes meeting high likelihood of OUD criteria upon chart review.

RESULTS

Only 57.7 % of patients were categorized as high likelihood of OUD; 16.5 % were categorized as limited aberrant opioid use, 18.9 % prescribed opioid use without evidence of aberrant use, and 6.9 % insufficient information.

CONCLUSIONS

Patients assigned ICD codes for opioid use, dependence, or abuse often lack documentation of meeting OUD criteria. Many receive long-term opioid therapy for chronic pain without evidence of misuse. Robust methods of identifying individuals with OUD are crucial to improving access to clinically appropriate treatment.

Information on doctor and pharmacy shopping for opioids adds little to the identification of presumptive opioid abuse disorders in health insurance claims data

Background

Doctor and pharmacy shopping (“Shopping”) for opioids is related to opioid abuse and is associated with opioid overdose and death. Lacking identifiers for prescribers and pharmacies, many data resources (notably the US FDA’s Sentinel System) cannot evaluate Shopping. We used data in which presumptive Shopping could be identified. We investigated whether US health insurance claims data could perform as well as Shopping to identify people with evidence for opioid abuse.

Methods

In this cross-sectional study, we examined health insurance claims from 164,923 persons with at least two dispensing of opioids in 18 months, the first occurring in 2012. Evidence for the presence of a possible opioid abuse disorder was drawn from predictive patterns of drug fills, diagnoses and care-seeking identified in a companion research project, and Shopping was determined using a published index. The prevalence of presumptive opioid abuse was examined across levels of Shopping. The comparison between Shopping and insurance-claims-derived covariates in the detection of apparent opioid abuse was examined in multiple regression analyses.

Results

Despite a strong correlation between presumptive opioid abuse and Shopping, most persons with extensive Shopping did not manifest presumptive opioid abuse, and half of the population with presumptive opioid abuse did not exhibit Shopping. As Shopping ranged from “None” to “Extensive,” the prevalence of presumptive opioid abuse increased from 0.28 to 5.0 per 100. The discriminating power of Shopping for identifying opioid abuse could be replaced using insurance claims data.

Conclusion

The results suggest that patient characteristics that can be inferred from insurance claims data provide as complete discrimination of persons with presumptive opioid abuse as does a full assessment of doctor and pharmacy shopping. The inference rests on patterns of health services and drug dispensing that are indicative of doctor–pharmacy shopping and of opioid abuse. There was no direct evaluation of patients. The extent to which the conclusions are generalizable beyond the study population – Americans with health insurance coverage in the early part of this decade – is uncertain in a quantitative sense. The qualitative conclusion is that diagnostic data in health insurance databases can be predictive of behaviors consistent with opioid abuse and that more elaborate indices such as doctor and pharmacy shopping may add little.

Registration: ClinicalTrials.gov study number: NCT02668549.

Trends of Opioid Use Disorder Among Hospitalized Patients With Chronic Pain

BACKGROUND

Patients with chronic pain treated with opioids are at an increased risk for opioid misuse or opioid use disorder (OUD). Recent years have seen a stark increase in abuse, misuse, and diversion of prescription opioid medications. The aim of this study was to investigate trends in changing rates of opioid use disorder among patients with chronic pain.

METHODS

The National Inpatient Sample (NIS) database identified chronic pain admissions with OUD from 2011 to 2015. Patients were identified from the NIS database using International Classification of Diseases (9th and 10th revisions) diagnosis codes for chronic pain and OUD. Annual estimates and trends were determined for OUD, patient characteristics, OUD among subgroups of chronic pain conditions, and discharge diagnosis.

RESULTS

We identified 10.3 million patients with chronic pain. Of these, 680,631 patients were diagnosed with OUD. The number of patients with OUD increased from 109,222 in 2011 to 172,680 in 2015 (P < 0.001). Similarly, there were upward trends of OUD among females (53.2% to 54.5%; P = 0.09), patients 65 to 84 years of age (11.8% to 17%; P < 0.001), Medicare-insured patients (39.5% to 46.0%; P < 0.01), patients with low annual household incomes (27.8% to 33.3%; P < 0.001), and patients with cannabinoid use disorder (7.2% to 8.3%; P = 0.01). The prevalence of OUD increased from 2011 to 2015 in patients with chronic regional pain syndrome (5.53% to 7.46%; P = 0.01) and spondylosis (1.32% to 1.81%; P < 0.001).

CONCLUSIONS

These findings suggest that the prevalence of OUD increased substantially from 2011 to 2015. Disparities of OUD with increasing opioid use among vulnerable populations including women, those with Medicare insurance, tobacco use disorder, and low annual income should be explored further.

Evidence for the use of cannabinoids in Parkinson's disease

Cannabis and synthetic cannabinoid formulations have now been legally approved in several countries for treatment of patients with Parkinson's disease (PD). Hence, PD patients consult physicians more frequently for prescription of cannabinoids to alleviate symptoms that might not respond well to dopaminergic treatment. Despite the increasing volume of research generated in the field of cannabinoids and their effect on Parkinson's disease, there is still paucity of sufficient clinical data about the efficacy and safety in PD patients. There is increasing understanding of the endocannabinoid system, and the distribution of cannabinoid receptors in basal ganglia structures might suggest potential benefit on parkinsonian symptoms. Concerning clinical research, only one of to date four conducted randomized placebo-controlled trials showed an effect on motor symptoms with alleviation of levodopa-induced dyskinesia. There are a growing number of uncontrolled trials and case reports that suggest beneficial effects of cannabinoids in PD patients. However, the variety of substances investigated, the varying routes of intake, differing doses and time courses make it difficult to compare data. We here provide an overview of the current literature in this field and discuss a pragmatic approach for the clinical use of cannabinoids in PD.

The Prevalence and Cost of Medicare Beneficiaries Diagnosed and At Risk for Opioid Abuse, Dependence, and Poisoning

BACKGROUND

Reliance on prescription opioids to manage pain has been associated with increases in diversion, overdose, and addiction. Prevalence of misuse and abuse has been shown to be higher among government-insured populations than commercially insured populations. However, the prevalence and costs of misuse/abuse among the Medicare fee-for-service (FFS) population has not been studied.

OBJECTIVES

To (a) determine the prevalence and costs of prescription opioid misuse/abuse and (b) evaluate the prevalence and costs associated with those identified as at risk for opioid misuse/abuse in Medicare FFS beneficiaries.

METHODS

This retrospective case-control study used Medicare claims data for the calendar years of 2010 and 2011 and included Medicare beneficiaries aged at least 18 years. The index date was the date of first diagnosed misuse/abuse or at risk for abuse and had to occur between July 1, 2010, and June 30, 2011, and beneficiaries had to have at least 6 months continuous eligibility before and after the index date. Matching (1:1) was used for comparing opioid misusers/abusers with nonabuser controls, as well as comparing patients at risk for opioid abuse with controls not at risk for abuse. Controls were matched to cases by gender, age, disability, and geographic region. The index date of the control patient was set equal to the index date of the matched case.

RESULTS

Prevalence of misuse/abuse in the Medicare FFS population was 13.1 per 1,000 persons, with the majority among patients receiving Medicare based on disability (76.2%). The prevalence of at risk for misuse/abuse was 117.4 per 1,000 persons. Approximately half of the Medicare FFS patients used an opioid. Overall total annual unadjusted mean costs of health care resources were significantly greater for abusers than for matched controls ($46,194 vs. $21,964; P < 0.0001), with a mean annual excess cost of $24,230. The overall total adjusted 6-month post-index mean costs of health care resources for abusers was significantly greater than that of matched controls ($33,942 vs. $10,754; P < 0.0001), with a mean excess cost of $23,188.

CONCLUSIONS

The prevalence of diagnosed abuse among Medicare FFS population (13.1 per 1,000 persons) was higher than other payer groups studied using similar ICD-9-CM codes, and the majority of abuse was among those receiving Medicare based on disability (76.2%). The prevalence of at-risk abuse was 9 times higher than the prevalence of diagnosed abuse. As with other studies, health care resource utilization and costs were significantly greater for diagnosed abuse than matched controls.

DISCLOSURES

This study was sponsored by Pfizer. Roland is a Pfizer employee and stockholder and was involved in all aspects of the study as part of a mid-career fellowship in pharmacoeconomics with the University of Utah. Ye and Stevens are employees of University of Utah, and Oderda was an employee of University of Utah, which received financial support from Pfizer in connection with the development of this manuscript. Oderda also reports consulting fees from Pfizer, Trevena, and Pacira, unrelated to this study. The results of this study were presented at the Academy of Managed Care Pharmacy Nexus 2015; October 26-29, 2015; Orlando, FL, and the AMCP Managed Care & Specialty Pharmacy Annual Meeting 2016; April 19-22, 2016; San Francisco, CA.

Opioid tolerance and clinically recognized opioid poisoning among patients prescribed extended-release long-acting opioids

BACKGROUND

In recognition of potential for increased overdose risk, drug labels for extended-release and long-acting (ER/LA) opioids emphasize the need for established opioid tolerance prior to initiating high dosages.

OBJECTIVES

Describe the proportion of patients with opioid tolerance prior to initiation of 90 morphine milligram equivalents (MME) ER/LA opioids and examine subsequent risk of opioid poisoning.

METHODS

We used Truven Health Analytics' MarketScan Databases (2006-2015) to identify patients initiating ER/LA opioids ≥90 MME. We examined prescription histories and describe the proportion of initiators with opioid tolerance (defined as ≥7 days of 60 MME in the prior 14 days). We adjusted for age, sex, year of initiation, and baseline comorbidities using inverse probability of treatment weighted Cox proportional hazards models. We estimated adjusted hazard ratios and 95% confidence intervals for the effect of opioid tolerance on the risk of clinically recognized opioid poisoning (based on diagnosis codes) in specific periods (0-7, 8-30, 31-90, and 91-365 days) following initiation.

RESULTS

Among 372 038 initiators, 38% did not meet opioid tolerance criteria. The proportion of nontolerant initiators was highest among those initiating methadone (44%) and fentanyl (42%). Nontolerant patients were 37% more likely to be diagnosed with opioid poisoning (adjusted hazard ratios = 1.37 [1.07, 1.76]) in the week following ER/LA initiation.

CONCLUSIONS

Over one-third of patients initiating ≥90 MME ER/LA opioids did not have evidence of opioid tolerance. The 7 days following high dose ER/LA initiation may represent a high-risk period for clinically diagnosed opioid poisoning in patients who do not have prior opioid tolerance.

Drug use generations and patterns of injection drug use: Birth cohort differences among people who inject drugs in Los Angeles and San Francisco, California

OBJECTIVES

A robust literature documents generational trends in drug use. We examined the implications of changing national drug use patterns on drug injection histories of diverse people who inject drugs (PWID).

METHODS

Drug use histories were collected from 776 active PWID in 2011-13. Using descriptive statistics, we examine drug use initiation by year and birth cohort (BC) differences in drug first injected. A multivariate linear regression model of time to injection initiation ([TTII] (year of first injection minus year of first illicit drug use) was developed to explore BC differences.

RESULTS

The first drug injected by BC changed in tandem with national drug use trends with heroin declining from 77% for the pre-1960's BC to 58% for the 1960's BC before increasing to 71% for the 1990's BC. Multivariate linear regression modeling found that shorter TTII was associated with the 1980's/1990's BC (-3.50 years; 95% Confidence Interval [CI]=-0.79, -6.21) as compared to the 1970's BC. Longer TTII was associated with being female (1.65 years; 95% CI=0.40, 2.90), African American (1.69 years; 95% CI=0.43, 2.95), any substance use treatment prior to injection (4.22 years; 95% CI=2.65, 5.79), and prior non-injection use of drug that was first injected (3.29 years; 95% CI=2.19, 4.40).

CONCLUSION

National drug trends appear to influence injection drug use patterns. The prescription opiate drug era is associated with shorter TTII. Culturally competent, demographically and generationally-targeted prevention strategies to combat transitions to drug injection are needed to prevent or shorten upstream increases in risky drug use practices on a national level.