Fentanyl Quality Assurance Project Prompted Change in Clinical Workflow and Test Configurations

Abstract

Background

Deaths attributable to fentanyl (FEN, a synthetic opioid) are high in Appalachia and highest in West Virginia. The goal of the study was to determine FEN prevalence among specimens submitted for definitive opioid testing and monitor responses to provider notifications of unexpected FEN findings during Q1 2020.

Methods

All definitive opioid test data were reviewed daily for FEN signatures in Q1 2020. Unexpected FEN results were communicated to providers and monitored for 10 days to record actions taken. Prevalence data were categorized. Behavioral Medicine (BMED) leaders analyzed January data and implemented FEN screening in the clinic. BMED Q1 clinic visits and order volumes for drug screens were reviewed after Q1.

Results

FEN positivity was 11% in Q1; >60% of findings were unexpected. Actions were taken for 54% of notifications in January but only 18% in March. Notifications required 70 hours of combined laboratory effort each month. BMED providers ordered 44% of definitive opioid tests and 69% of definitive FEN tests. Data prompted the addition of FEN to routine drug screen panels in the laboratory, and a 10% random FEN screening rate in the BMED opioid use disorder clinics (COAT).

Conclusions

Prevalence of FEN positivity was higher than initially expected, even for this region in Appalachia. Expanded presence of FEN screening should assist BMED providers with clinical efforts and help identify patients in need of intervention/therapy.

Background

The opioid crisis in the United States has shown steady increases in age-adjusted deaths each year; Appalachia (especially West Virginia, WV) has been hardest hit owing to a combination of economic, industrial, and geographic influences (1–6). WV has had the highest number of age-adjusted opioid-related deaths each year since 2013 (2), and of the 49.6 opioid-related deaths per 100 000 persons in WV in 2017, 74% (618 of 833) were attributed to synthetic opioids such as fentanyl (FEN) (1, 3). Particularly in WV, the opioid crisis has represented one symptom of a larger disorder of healthcare and addiction treatment access disparities, coupled with health declines (e.g., obesity, diabetes) that hasten death (4–6).

FEN and its analogs have higher potency than heroin and opiates such as morphine and oxycodone, and a higher risk of morbidity and mortality associated with misuse. Regionally high prevalence of opioid misuse could include intentional and unintentional FEN exposure. Given the broad range of providers who encounter patients with substance use disorders (SUD), local testing for FEN should be considered in high-prevalence areas. Also, given clinical laboratories’ strength in generating clinical data, they have a prime opportunity for estimating prevalence of drug positivity among local patients, and informing local providers. Providers and laboratories should thus work together to determine needed drug test offerings, and track regional drug use trends.

Despite the figures and educational efforts directed at providers, definitive testing for misused drugs only followed presumptive positive screening at the West Virginia University Hospital (WVUH) 25% of the time in 2019 (Table 1 in the online Supplemental Data). Further, tests in local routine drug screen panels lack cross-reactivity with FEN, and FEN was not included in the testing menu. Also, until now, there have been no waived point-of-care FEN screening tests approved by the FDA. These devices are used in WVUH Comprehensive Opioid Addiction Treatment (COAT) and Pain Management clinics, and administered by personnel who do not meet state qualifications for moderate-complexity testing. Thus, the absence of rapid FEN testing put our Behavioral Medicine providers (BMED) at a relative loss.

Opioid-related death and morbidity trends in WV, and a growing collaboration between the laboratory and BMED, prompted initiatives to implement local expansions of opioid screening and confirmatory methodology. An FDA-approved automated FEN screening test, and an in-house definitive opioid test panel in the clinical mass spectrometry laboratory, were launched in 2019. In a prelaunch series of “practice runs” (i.e., pilots) for developing the daily definitive opioid testing workflow, randomly selected specimen remnants were positive for FEN >10% of the time. Thus, a Q1 2020 quality assurance (QA) study was designed to include daily checks for FEN-positive specimens, with subsequent notifications to providers when FEN was unexpected. The hypotheses were: (a) that the presence of FEN in urine specimens submitted for definitive opioid testing was higher than assumed rates (based on historical ordering trends); and (b) that new measures to ensure adequate testing coverage and clinical action would be necessary, both in the laboratory and clinic settings.

Methods and Materials

Study Approval

This study was performed under the auspices of a WVUH Laboratory-submitted and -maintained IRB approval (1306049573).

Laboratory Workflow for Definitive Testing

Urine specimens were submitted to the WVUH clinical mass spectrometry core for definitive opioid testing; testing was completed as ordered by the clinician. Using an unpublished, locally developed method, aliquots of urine were hydrolyzed to remove glucuronide conjugates, spiked with internal standards, and diluted for injection on the LC-MS/MS system. Components measured included 21 opioids and primary metabolites; definitive testing could be accessed via reflex from positive screens for available tests, or directly ordered by the provider. Cutoffs were 0.2 ng/mL for FEN and 2 ng/mL for norfentanyl. Reporting for any orderable covered by the panel provided: 4 adulteration test results (creatinine, specific gravity, pH, and oxidant check); qualitative results for negative components; and quantitative results for positive components.

Workflow for Notification System

During postrun data review and reporting each weekday, bench scientists identified positive FEN signatures among specimens tested; information shared with the Director included specimen identification number, FEN and norFEN results, and medical record number (Fig. 1). The Director accessed electronic health records (Epic, August 2019 version) in compliance with QA and study protocol to determine if the presence of FEN was expected (i.e., medically administered for pain control) or not. Unexpected FEN was reported to the listed authorizing provider via Epic inbox using a form letter (Supplemental Table 2).

The Director monitored the notification for up to 10 days. If the provider acknowledged the message with a reply, sent a note to another provider related to the notification and/or ordered any FEN screen or confirmatory test (i.e., on the current specimen or a future one), this was considered action taken for the purposes of the QA program. If the provider did not act per these parameters by the date given, the notification was marked as “Expired” and monitoring ceased. On February 18, summary data to-date were reviewed by the Director and BMED leads; data capture for Q1 was completed.

COAT Clinic Workflow

Patients in the COAT clinic were routinely screened for illicit substance use and compliance with treatment. Waived screening cups containing standard urine drug screen and adulteration check strips composed the first stage of screening and were performed on ≥50% of all COAT visits (Supplemental Data, COAT-Volumes tab). Frequency of screening varied by patient, averaging weekly for patients in early recovery, and less frequently for patients who accrued more sober time (data not shown).

Unexpected positive and/or negative screens are routed to the central laboratory for definitive testing. In the COAT clinic, patients were asked about any relapses, and whether they have been compliant with buprenorphine treatment. Examples of instances prompting definitive testing included: (a) a patient indicating sobriety who had a positive screening test for opiates; and (b) concern for simulated compliance. The intents of the approach were promoting patient safety, and facilitating patient recovery. For example, confirmed discrepancies between patient reports and definitive tests could escalate patients to more frequent clinic visits, or prompt referral to inpatient treatment facilities.

For drugs not covered by screening cups, definitive tests were ordered by the providers via the central laboratory (e.g., ethanol biomarkers, gabapentin). Each test was ordered on approximately 10% of COAT visitors in Q1 2020, and largely performed randomly among patients. However, targeted testing on patients with a known history of misuse also occurred. In January of 2020, there was no screening target for FEN in COAT clinic.

Additional Quality and Clinical Data Collections

Laboratory-based QA reports (Epic) were configured to provide counts of all specimens and tests submitted to the mass spectrometry core for definitive opioid testing in Q1 2020. Counts of all specimens submitted for drug screening tests in the WVUH Laboratory were also captured to give scope to the potential problem, beyond the definitive testing cohort; those reports were counted according to drug test or panel ordered. (The online Supplemental Data file contains tabs for each report type and month.)

COAT clinic visits, breakouts by first-time and return patients, and breakouts by drug screens and definitive tests ordered—FEN, gabapentin, ethanol metabolites, and point-of-care urine drug screen cups—were compiled for Q1 by BMED from existing Epic reports. Counts were expressed as summary data by weeks.

Data Analysis

Data in each FEN monitoring tab (Supplemental Data) included: FEN and norfentanyl results; day, time, and location the specimen was created; definitive test(s) ordered; whether or not the provider was notified (yes/no); reason NOT notifying as applicable; day and time of notification; binary field for action (yes/no); date and time action taken (also used for date of expiry to aid monitoring); what action was taken or expiry; comments; and time to action in hours (= action day/time − notification day/time). Screening and definitive test QA summary sheets—organized by month—yielded patient age and gender, test ordered, component results, and ordering location.

Calculations included counts, percentages, medians, and percentiles (2.5^th and 97.5^th) performed in Microsoft Excel 2016. No calculations of statistical significance were conducted on the notification study data and volume data, due to low numbers of notifications and changing volume in Q1 coincident with SARS-CoV-2-related lockdowns.

Results

FEN-Positive Specimens and Notifications

The Director reviewed the most FEN-positive specimens in January, and the least in March. Relative to specimen volume tested in the mass spectrometry laboratory, however, notification volumes were stable and averaged 11% (Table 1). Pain control with FEN on inpatients accounted for most reasons for not notifying providers during the study. Of notifications sent, providers acted more often in January; declining numbers of actions taken were noted in February and March. Conversely, unprompted orders for definitive FEN testing increased throughout Q1. Expired notifications also increased throughout Q1.

Laboratory Resource Demands during the QA Study

The notification process required an average of 0.5 hours of bench scientist work, and 3 hours of Director work, per weekday in Q1 (Fig. 1). This work amounted to 70 hours per month; in full-time equivalents, this amounted to 0.125 bench scientists, and 0.75 Directors.

Definitive Test Orders by Provider/Service Type

BMED ordered the majority of definitive opioid tests, definitive FEN tests, and unprompted FEN orders in Q1 2020 (Table 2).

Clinical Actions Taken Based on QA Data

The high rate of unexpected FEN found in January was the focus of an interim QA project review between the Director and BMED leaders in February. A FEN screening target of 10% was promptly implemented for COAT clinic patients by the BMED leaders—similar to existing random screening rates for ethanol biomarkers and gabapentin. The implementation was viewed as successful and relatively rapid; while 3 FEN test orders were placed in 906 (0.3%) COAT visits in January, 40 were placed in 657 (6%) COAT visits in March. Of 70 FEN screens performed in COAT clinic in Q1 2020, 84% were performed on patients new to the clinic, 13% were targeted screens of high-risk patients, and 3% were truly random. (These data can be found in Supplemental Data, COAT-volumes tab.)

Discussion

Drug testing is a crucial part of monitoring the recovery of patients with SUD; it is also important in directing these patients to appropriate levels of care to achieve sustained recovery and prevent overdose deaths. FEN and its analogs have become the primary drivers of overdose deaths in recent years in WV (1–3, 7), and as such it is vital to include screening for FEN in any laboratory-based drug screening approach in this region.

A strength of this study is the analysis of a continuous data set collected at WVUH in Q1 2020. The data demonstrate an 11% overall FEN positivity rate among specimens submitted for definitive testing, and that most definitive FEN testing orders came from BMED. Considering that this study sample represents only a subset of patients undergoing drug screening at WVUH, the QA study exposes several gaps in the reach of available laboratory test offerings in this region. Exposing such gaps proves useful in making changes, because the demonstration of need is clear in this case.

Another strength of the study is the focused nature of the notification study, and deliberate monitoring of actions taken for a dedicated 90-day time interval. This approach allowed the laboratory to assess the mix of patients more commonly associated with the definitive testing, monitor QA for the newly launched method, and identify frequent test-users among providers. The data herein suggest that extending FEN screening to routine automated drug screens would capture more patients taking FEN in general, inform providers automatically/without interruption, and enable better capture for followup clinical and QA monitoring efforts. Also, the study was successful in identifying clinical champions to support this change. As a result, effective May 1, 2020, WVUH drug screen panels—reflexive, and nonreflexive—include FEN.

The FEN screen at WVUH has a cutoff of 1 ng/mL and detects FEN, norfentanyl, butyryl-fentanyl, and acetyl-fentanyl (8). However, the typical caveats of immunoassay-based screening tests apply, and further study relating to the quality of these universalized screening results should be conducted. Preliminary data from June and July 2020 demonstrate that 127 of 198 (64%, data not shown) definitive FEN orders were positive—an early sign that adding FEN to routine screening panels may help identify patients in need of intervention/referral to COAT clinic. Also, efforts are underway to evaluate medications lists in FEN-negative definitive tests. The goal of this review is to see if a pattern of prescribed medications potentially interfering with the screening test can be identified, tested, and shared with providers. Since more assessment is needed, a followup project is being planned.

Adding FEN to routine drug screens should further efforts to detect and/or monitor opioid misuse in emergency department, inpatient, and outpatient settings in this region of Appalachia (9). This approach universalizes the search for relevant opioids among patients presenting throughout the care structure, to encourage earlier clinical capture and intervention. BMED providers commonly acquire patients by referral from other services; upfront knowledge of potential drugs misused can assist with early intervention and directed treatment approaches.

A limitation of this study was a lack of data-gathering among providers regarding reasons for expired notifications. However, since alert fatigue and perceived irrelevance of alerts in modern electronic health records are listed among hindrances to clinical work among providers (10–13), further questionnaires and messages were avoided in the study design. Only the notification form letter was sent, with no reminders. However, despite the lack of later/followup inquiries with providers about reasons for letting notifications expire, the study data reinforce these already-reported concepts. We show a relatively rapid decline in action taken and concomitant rise in expirations within the quarter studied; since the providers involved changed minimally and a form letter was used, the system was very similar to existing and published automated alerts, and associated providers relatively constant.

The expansion of FEN screening to outpatient clinics still presents challenges. Because of lacking rapid/waived drug screening options for FEN, an inherent resulting delay associated with central (screening) and specialty (definitive) laboratory testing can be expected. Given that the implementation of random FEN screening in COAT clinic yielded a 70% positivity rate, isolated to the second half of Q1 2020, this study confirms that the practice successfully identifies surreptitious users. Also, the data show that in the COAT clinic, this expansion is balanced by a relatively small cost of false-positive screens.

Another point of further exploration involves FEN analogs, which are not currently included in the WVUH definitive test. Since butyryl-fentanyl and acetyl-fentanyl can be detected with the screening test, a pilot study of a FEN definitive panel expansion to include these analogs could be considered.

Conclusions

FEN positivity in this region of WV is higher than previous estimates. Efforts to increase screening should aid the identification of patients needing intervention and therapy. FEN testing in the COAT setting should continue and perhaps expand.

Supplemental Material

Supplemental material is available at The Journal of Applied Laboratory Medicine online.

Author Contributions: All authors confirmed they have contributed to the intellectual content of this paper and have met the following 4 requirements: (a) significant contributions to the conception and design, acquisition of data, or analysis and interpretation of data; (b) drafting or revising the article for intellectual content; (c) final approval of the published article; and (d) agreement to be accountable for all aspects of the article thus ensuring that questions related to the accuracy or integrity of any part of the article are appropriately investigated and resolved.

D.H. Tacker, statistical analysis, administrative support; J.H. Berry, provision of study material or patients.

Authors' Disclosures or Potential Conflicts of Interest: No authors declared any potential conflicts of interest.

Role of Sponsor: No sponsor was declared.

Acknowledgments: The authors thank (in alphabetic order) the WVUH mass spectrometry clinical laboratory scientists (Julie Domico, Lois Frazee, Carole Mahaffey, Meranda Mancino, and Joyce Topardo). We thank these individuals, for not only their everyday efforts in the mass spectrometry section, but also their additional assistance in this QA study.

REFERENCES