Development and evaluation of a text analytics algorithm for automated application of national COVID-19 shielding criteria in rheumatology patients

INTRODUCTION

At the beginning of the COVID-19 pandemic, the UK's Scientific Committee issued extreme social distancing measures, termed 'shielding', aimed at a subpopulation deemed extremely clinically vulnerable to infection. National guidance for risk stratification was based on patients' age, comorbidities and immunosuppressive therapies, including biologics that are not captured in primary care records. This process required considerable clinician time to manually review outpatient letters. Our aim was to develop and evaluate an automated shielding algorithm by text-mining outpatient letter diagnoses and medications, reducing the need for future manual review.

METHODS

Rheumatology outpatient letters from a large UK foundation trust were retrieved. Free-text diagnoses were processed using Intelligent Medical Objects software (Concept Tagger), which used interface terminology for each condition mapped to Systematized Medical Nomenclature for Medicine-Clinical Terminology (SNOMED-CT) codes. We developed the Medication Concept Recognition tool (Named Entity Recognition) to retrieve medications' type, dose, duration and status (active/past) at the time of the letter. Age, diagnosis and medication variables were then combined to calculate a shielding score based on the most recent letter. The algorithm's performance was evaluated using clinical review as the gold standard. The time taken to deploy the developed algorithm on a larger patient subset was measured.

RESULTS

In total, 5942 free-text diagnoses were extracted and mapped to SNOMED-CT, with 13 665 free-text medications (n=803 patients). The automated algorithm demonstrated a sensitivity of 80% (95% CI: 75%, 85%) and specificity of 92% (95% CI: 90%, 94%). Positive likelihood ratio was 10 (95% CI: 8, 14), negative likelihood ratio was 0.21 (95% CI: 0.16, 0.28) and F1 score was 0.81. Evaluation of mismatches revealed that the algorithm performed correctly against the gold standard in most cases. The developed algorithm was then deployed on records from an additional 15 865 patients, which took 18 hours for data extraction and 1 hour to deploy.

DISCUSSION

An automated algorithm for risk stratification has several advantages including reducing clinician time for manual review to allow more time for direct care, improving efficiency and increasing transparency in individual patient communication. It has the potential to be adapted for future public health initiatives that require prompt automated review of hospital outpatient letters.

Trends for opioid prescribing and the impact of the COVID-19 pandemic in patients with rheumatic and musculoskeletal diseases between 2006 and 2021

OBJECTIVE

To investigate opioid prescribing trends and assess the impact of the COVID-19 pandemic on opioid prescribing in rheumatic and musculoskeletal diseases (RMDs).

METHODS

Adult patients with RA, PsA, axial spondyloarthritis (AxSpA), SLE, OA and FM with opioid prescriptions between 1 January 2006 and 31 August 2021 without cancer in UK primary care were included. Age- and gender-standardized yearly rates of new and prevalent opioid users were calculated between 2006 and 2021. For prevalent users, monthly measures of mean morphine milligram equivalents (MME)/day were calculated between 2006 and 2021. To assess the impact of the pandemic, we fitted regression models to the monthly number of prevalent opioid users between January 2015 and August 2021. The time coefficient reflects the trend pre-pandemic and the interaction term coefficient represents the change in the trend during the pandemic.

RESULTS

The study included 1 313 519 RMD patients. New opioid users for RA, PsA and FM increased from 2.6, 1.0 and 3.4/10 000 persons in 2006 to 4.5, 1.8 and 8.7, respectively, in 2018 or 2019. This was followed by a fall to 2.4, 1.2 and 5.9, respectively, in 2021. Prevalent opioid users for all RMDs increased from 2006 but plateaued or dropped beyond 2018, with a 4.5-fold increase in FM between 2006 and 2021. In this period, MME/day increased for all RMDs, with the highest for FM (≥35). During COVID-19 lockdowns, RA, PsA and FM showed significant changes in the trend of prevalent opioid users. The trend for FM increased pre-pandemic and started decreasing during the pandemic.

CONCLUSION

The plateauing or decreasing trend of opioid users for RMDs after 2018 may reflect the efforts to tackle rising opioid prescribing in the UK. The pandemic led to fewer people on opioids for most RMDs, providing reassurance that there was no sudden increase in opioid prescribing during the pandemic.

Psoriatic arthritis: the role of self-reported non-adherence, non-trough drug levels, immunogenicity and conventional synthetic DMARD co-therapy in adalimumab and etanercept response

Objective

The aim of this study was to assess the relationship between self-reported non-adherence, non-trough drug levels, immunogenicity and conventional synthetic DMARD (csDMARD) co-therapy in TNF inhibitor (TNF-i) drug response in PsA.

Methods

Serum samples and adherence questionnaires were collected at baseline, 3, 6 and 12 months for PsA patients prescribed TNF-i. Non-trough adalimumab (ADL) and etanercept (ETN) drug levels were measured at 3 and 6 months using commercially available ELISAs. Clinical response was assessed using PsA response criteria (PsARC) and change in 28-joint DAS (ΔDAS28) between baseline and 3, 6 and 12 months.

Results

In 244 PsA patients (52.5% ADL and 47.5% ETN), self-reported non-adherence was associated with PsARC non-response over 12 months using generalized estimating equation (GEE) modelling (P = 0.037). However, there was no significant difference between non-trough ADL or ETN drug levels based on self-reported non-adherence. Higher ETN levels at 3 months were associated with PsARC response at 3 (P = 0.015), 6 (P = 0.037) and 12 months (P = 0.015) and over 12 months using GEE modelling (P = 0.026). Increased ADL drug levels at 3 months were associated with greater ΔDAS28 at 3 months (P = 0.019). ADL anti-drug antibody-positive status was significantly associated with lower 3- and 6-month ADL levels (P < 0.001) and ΔDAS28 and PsARC response at 3, 6 and 12 months. Meanwhile, MTX co-therapy was associated with a reduction in immunogenicity at 3 and 6 months (P = 0.008 and P = 0.024).

Conclusion

Although both were associated with reduced response, the objectively measured non-trough drug levels showed more significant associations with drug response than self-reported non-adherence measures.

"Take up to eight tablets per day": Incorporating free-text medication instructions into a transparent and reproducible process for preparing drug exposure data for pharmacoepidemiology

PURPOSE

Routinely collected prescription data provides drug exposure information for pharmacoepidemiology, informing start/stop dates and dosage. Prescribing information includes structured data and unstructured free-text instructions, which can include inherent variability, such as "one to two tablets up to four times a day". Preparing drug exposure data from raw prescriptions to a research ready dataset is rarely fully reported, yet assumptions have considerable implications for pharmacoepidemiology. This may have bigger consequences for "pro re nata" (PRN) drugs. Our aim was, using a worked example of opioids and fracture risk, to examine the impact of incorporating narrative prescribing instructions and subsequent drug preparation assumptions on adverse event rates.

METHODS

R-packages for extracting free-text medication prescription instructions in a structured form (doseminer) and an algorithm for transparently processing drug exposure information (drugprepr) were developed. Clinical Practice Research Datalink GOLD was used to define a cohort of adult new opioid users without prior cancer. A retrospective cohort study was performed using data between January 1, 2017 and July 31, 2018. We tested the impact of varying drug preparation assumptions by estimating the risk of opioids on fracture risk using Cox proportional hazards models.

RESULTS

During the study window, 60 394 patients were identified with 190 754 opioid prescriptions. Free-text prescribing instruction variability, where there was flexibility in the number of tablets to be administered, was present in 42% prescriptions. Variations in the decisions made during preparing raw data for analysis led to marked differences impacting the event number (n = 303-415) and person years of drug exposure (5619-9832). The distribution of hazard ratios as a function of the decisions ranged from 2.71 (95% CI: 2.31, 3.18) to 3.24 (2.76, 3.82).

CONCLUSIONS

Assumptions made during the drug preparation process, especially for those with variability in prescription instructions, can impact results of subsequent risk estimates. The developed R packages can improve transparency related to drug preparation assumptions, in line with best practice advocated by international pharmacoepidemiology guidelines.

EULAR points to consider for therapeutic drug monitoring of biopharmaceuticals in inflammatory rheumatic and musculoskeletal diseases

OBJECTIVE

To develop EULAR points-to-consider for therapeutic drug monitoring (TDM) of biopharmaceuticals in inflammatory rheumatic and musculoskeletal diseases (RMDs).

METHODS

The points-to-consider were developed in accordance with EULAR standardised operation procedures by a multidisciplinary task force from eight European countries, based on a systematic literature review and expert consensus. Level of evidence and strength of the points-to-consider were determined, and mean levels of agreement among the task force were calculated using a 10-point rating scale.

RESULTS

Six overarching principles and 13 points-to-consider were formulated. The level of agreement among the task force for the overarching principles and points-to-consider ranged from 8.4 to 9.9.The overarching principles define TDM and its subtypes, and reinforce the underlying pharmacokinetic/pharmacodynamic principles, which are relevant to all biopharmaceutical classes. The points-to-consider highlight the clinical utility of the measurement and interpretation of biopharmaceutical blood concentrations and antidrug antibodies in specific clinical scenarios, including factors that influence these parameters. In general, proactive use of TDM is not recommended but reactive TDM could be considered in certain clinical situations. An important factor limiting wider adoption of TDM is the lack of both high quality trials addressing effectiveness and safety of TDM and robust economic evaluation in patients with RMDs. Future research should focus on providing this evidence, as well as on further understanding of pharmacokinetic and pharmacodynamic characteristics of biopharmaceuticals.

CONCLUSION

These points-to-consider are evidence-based and consensus-based statements for the use of TDM of biopharmaceuticals in inflammatory RMDs, addressing the clinical utility of TDM.

Assessing Immunogenicity of Biologic Drugs in Inflammatory Joint Diseases: Progress Towards Personalized Medicine

Biologic drugs have greatly improved treatment outcomes of inflammatory joint diseases, but a substantial proportion of patients either do not respond to treatment or lose response over time. Drug immunogenicity, manifested as the formation of anti-drug antibodies (ADAb), constitute a significant clinical problem. Anti-drug antibodies influence the pharmacokinetics of the drug, are associated with reduced clinical efficacy, and an increased risk of adverse events such as infusion reactions. The prevalence of ADAb differs among drugs and diseases, and the detection of ADAb also depends on the assay format. Most data exist for the tumor necrosis factor-alpha inhibitors infliximab and adalimumab, with a frequency of ADAb that ranges from 10 to 60% across studies. Measurement of ADAb and serum drug concentrations, therapeutic drug monitoring, has been suggested as a strategy to optimize therapy with biologic drugs. Although the recent randomized clinical Norwegian Drug Monitoring (NOR-DRUM) trials show promise towards a personalized medicine prescribing approach by therapeutic drug monitoring, several challenges remain. A plethora of assay formats, with widely differing properties, is currently used for measuring ADAb. Comparing results between different assays and laboratories is difficult, which complicates the development of cut-offs necessary for guidelines and the implementation of ADAb measurements in clinical practice. With the possible exception of infliximab, limited data on clinical relevance and cost effectiveness exist to support therapeutic drug monitoring as a routine clinical strategy to monitor biologic drugs in inflammatory joint diseases. The aim of this review is to provide an overview of the characteristics and prevalence of ADAb, predisposing factors to ADAb formation, commonly used assessment methods, clinical consequences of ADAb, and the potential implications of ADAb assessments for everyday treatment of inflammatory joint diseases.

Non-Trough adalimumab and certolizumab drug levels associated with a therapeutic EULAR response in adherent patients with rheumatoid arthritis

OBJECTIVES

Interventions aimed at increasing tumour-necrosis factor-alpha inhibitor serum drug levels (SDLs) may improve treatment response; however, previous studies suggesting SDL cut-offs have not accounted for treatment adherence. The aim of this study was to establish the relationship between Adalimumab/Certolizumab SDLs and EULAR good vs non/moderate response, and to define SDL cut-offs associated with good response in fully adherent patients.

METHODS

In a prospective observational study, 475 patients with RA were treated with Certolizumab (n = 192) or Adalimumab (n = 283). At baseline, 3/6/12-months patients had DAS28, self-reported treatment adherence, and SDLs measured. Fully adherent patients were analysed as a subgroup. Follow-up data at 3/6/12-months was analysed separately. Median SDLs were compared in good vs non/moderate response, and receiver-operator characteristics curves (ROC) were used to establish cut-off SDLs.

RESULTS

Fully adherent good responders had significantly higher median Adalimumab/Certolizumab SDLs compared with non/moderate responders (p= 0·04 and p= 0·0005, respectively). ROC analysis reported 3-month non-trough Adalimumab SDLs discriminated good vs non/moderate response with an AUC of 0·63 (95% CI 0·52-0·75), with a cut-off of 7·5mg/l being 39·1% specific, and 80·9% sensitive. Similarly, 3-month non-trough Certolizumab SDLs discriminated good vs non/moderate response with an AUC of 0·65 (95% CI 0·51-0·78), with a cut-off of 26·0mg/l being 43·9% specific, and 77·8% sensitive.

CONCLUSION

In fully adherent patients, higher SDLs are detected in good responders suggesting that interventions to improve SDLs such as encouraging adherence could improve treatment response. 3-month non-trough SDL cut-offs of 7·5mg/l for Adalimumab and 26·0mg/l for Certolizumab may be useful in clinical practice.

Development and external validation of prediction models for adverse health outcomes in rheumatoid arthritis: A multinational real-world cohort analysis

BACKGROUND

Identification of rheumatoid arthritis (RA) patients at high risk of adverse health outcomes remains a major challenge. We aimed to develop and validate prediction models for a variety of adverse health outcomes in RA patients initiating first-line methotrexate (MTX) monotherapy.

METHODS

Data from 15 claims and electronic health record databases across 9 countries were used. Models were developed and internally validated on Optum® De-identified Clinformatics® Data Mart Database using L1-regularized logistic regression to estimate the risk of adverse health outcomes within 3 months (leukopenia, pancytopenia, infection), 2 years (myocardial infarction (MI) and stroke), and 5 years (cancers [colorectal, breast, uterine] after treatment initiation. Candidate predictors included demographic variables and past medical history. Models were externally validated on all other databases. Performance was assessed using the area under the receiver operator characteristic curve (AUC) and calibration plots.

FINDINGS

Models were developed and internally validated on 21,547 RA patients and externally validated on 131,928 RA patients. Models for serious infection (AUC: internal 0.74, external ranging from 0.62 to 0.83), MI (AUC: internal 0.76, external ranging from 0.56 to 0.82), and stroke (AUC: internal 0.77, external ranging from 0.63 to 0.95), showed good discrimination and adequate calibration. Models for the other outcomes showed modest internal discrimination (AUC < 0.65) and were not externally validated.

INTERPRETATION

We developed and validated prediction models for a variety of adverse health outcomes in RA patients initiating first-line MTX monotherapy. Final models for serious infection, MI, and stroke demonstrated good performance across multiple databases and can be studied for clinical use.

FUNDING

This activity under the European Health Data & Evidence Network (EHDEN) has received funding from the Innovative Medicines Initiative 2 Joint Undertaking under grant agreement No 806968. This Joint Undertaking receives support from the European Union's Horizon 2020 research and innovation programme and EFPIA.

POS1229 THE IMPACT OF COVID-19 ON MEDICATION NON-ADHERENCE IN A RHEUMATOID AND PSORIATIC ARTHRITIS UK COHORT

Background

In March 2020, as part of the UK’s COVID-19 prevention strategy, those identified as ‘clinically extremely vulnerable’, were advised to shield. This included a number of patients prescribed anti-rheumatic drugs, who were asked to continue their current treatment unless they developed symptoms of infection. Suboptimal treatment adherence (16.0%-81.0%) has been reported in patients with arthritic diseases, and is associated with psychological factors, including anxiety (1). Previous literature in non-UK cohorts has highlighted suboptimal adherence levels in immunosuppressed patients during the pandemic, although many were single centre studies (2,3).

Objectives

The aim of this multi-centre study is to investigate the impact of the COVID-19 pandemic on adherence to anti-rheumatic medications in patients with established rheumatoid (RA) and psoriatic (PsA) arthritis in the UK who had recently commenced a biologic or targeted synthetic DMARD.

Methods

Between September 2020 and May 2021, RA and PsA patients prescribed biologic or targeted synthetic anti-rheumatic drugs from two multi-centre observational studies (BRAGGSS and OUTPASS) were sent a questionnaire on medication usage, adherence, and perceptions to establish the impact of COVID-19 on these parameters. Patients were asked about compliance during the COVID-19 pandemic using a 5-point Likert scale (always, often, sometimes, rarely, and never) and the reason for non-adherence. Adherence was defined as never missing or delaying a dose, unless medically advised. Descriptive summary statistics were calculated, and logistic regression and Pearson’s chi-squared tests were employed to investigate variables associated with self-reported non-adherence.

Results

In total 159 questionnaires were returned (81.1% RA and 18.9% PsA). Methotrexate (53.5%) was the most frequently prescribed agent, followed by etanercept (25.2%), sulfasalazine (22.6%), hydroxychloroquine (21.4%) and adalimumab (19.5%). Furthermore, 68.6% of patients were prescribed ≥2 drugs. During the pandemic, 42.1% of patients reported missing or delaying a treatment dose for any reason. Adherence information was available for 97.5% of patients with 25.8% reporting non-adherence which was not medically advised. Methotrexate non-adherence was 27.1%, with similar levels reported for etanercept (20.0%), sulfasalazine (27.8%), hydroxychloroquine (35.3%) and adalimumab (29.0%). No drugs had significantly different adherence compared to methotrexate. Furthermore, there was no association between disease type or perception of disease control and adherence. Of non-adherent patients, 17.5% reported increased anxiety, fear, and increased risk due to the COVID-19 pandemic as an influencing factor. Meanwhile, 37.5% of non-adherent patients listed non-COVID-19 intentional reasons and 45.0% reported non-intentional reasons, with forgetting and running out of treatment listed most frequently.

Conclusion

In a UK cohort self-reported non-adherence was reported in 25.8% of patients during the COVID-19 pandemic, despite medical advice, with reasons including increased anxiety due to COVID-19.

References

[1]Medication adherence and persistence in patients with rheumatoid arthritis, psoriasis, and psoriatic arthritis: a systematic literature review. Patient Prefer Adherence. 2018;12:1483–503.

[2]Vakirlis E, Bakirtzi K, Papadimitriou I, Vrani F, Sideris N, Lallas A, et al. Treatment adherence in psoriatic patients during COVID-19 pandemic: Real-world data from a tertiary hospital in Greece. J Eur Acad Dermatology Venereol. 2020;34(11):e673–5.

[3]Polat Ekinci A, Pehlivan G, Gökalp MO. Surveillance of psoriatic patients on biologic treatment during the COVID-19 pandemic: A single-center experience. Dermatol Ther. 2020;(December 2020):19–22.

Acknowledgements

on behalf of the BRAGGSS consortium

Disclosure of Interests

Philippa Curry: None declared, Hector Chinoy Speakers bureau: UCB, Biogen, Consultant of: Novartis, Eli Lilly, Orphazyme, Astra Zeneca, Grant/research support from: Eli Lilly, UCB, Meghna Jani: None declared, Darren Plant: None declared, Kimme Hyrich Consultant of: consultancy/honoraria from AbbVie, Grant/research support from: Pfizer, BMS, Ann Morgan Speakers bureau: Roche, Chugai, Consultant of: GSK, Roche, Chugai, AstraZeneka, Regeneron, Sanofi, Vifor, Grant/research support from: Roche, Kiniksa Pharmaceuticals, Anthony G Wilson: None declared, John Isaacs Speakers bureau: Abbvie, Gilead, Roche, UCB, Grant/research support from: GSK, Janssen, Pfizer, Andrew Morris: None declared, Anne Barton Grant/research support from: I have received grant funding from Pfizer, Galapagos, Scipher Medicine and Bristol Myers Squibb., James Bluett Grant/research support from: Pfizer Limited. JB has received travel/conference fees from UCB, Pfizer and Eli Lilly

OP0051 TRENDS FOR OPIOID PRESCRIPTIONS AMONG PATIENTS WITH RHEUMATIC AND MUSCULOSKELETAL DISEASES BETWEEN 2006-2020

Background

Opioid prescribing has contributed to a North American epidemic with increasing trends in several European countries1. Rheumatic and musculoskeletal diseases (RMDs) are one of the most common indications for prescribed opioids despite there being little evidence on opioid prescribing and the benefit of long-term use in RMDs.

Objectives

To investigate national UK opioid prescribing trends by studying the patterns of opioid prescribing in new users with the following six RMDs: rheumatoid arthritis (RA), psoriatic arthritis (PsA), axial spondyloarthritis (AxSpA), systemic lupus erythematosus (SLE), osteoarthritis (OA) and fibromyalgia.

Methods

Patients aged 18 years and older with a diagnosis of RA, PsA, AxSpA, SLE and OA with a new episode of opioid use without cancer in the Clinical Practice Research Datalink (CPRD) were included between 01/01/2006 and 31/12/2020. CPRD is a database of anonymised UK primary care electronic health records representative of the national population. New opioid users were defined as individuals with RMDs who had a new episode of opioid use in a 2-year time window up to 6 months before or after an RMD diagnosis. Rates of new opioid users were calculated by dividing the number of new opioid users with an RMD per year by the number of eligible patients registered in CPRD per year. Age- and gender-standardised rates for new opioid users were obtained using direct standardisation for each RMD. Rates of opioid prescriptions among new users for each RMD were calculated by dividing the number of opioid prescriptions among new users with an RMD in the year they had new opioid episodes by patient-years of the new users with an RMD. Trends for the rates in the study period were tested using negative binomial regression. Significant change points were identified by looking at the points where the derivative (i.e. rate of change) of the trends for the rates crossed zero. Recurrent opioid users were defined as patients who had at least 3 opioid prescriptions issued within 3 months after a new opioid episode.

Results

This study included 21,505 RA patients, 8,392 PsA patients, 4,491 AxSpA patients, 4,508 SLE patients, 944,078 OA patients, and 33,829 fibromyalgia patients, who had new opioid episodes between 2006-2020. Whilst the overall trend for RA (2.7* vs 3.9*), PsA (1.0* vs 1.8*) and fibromyalgia (3.7* vs 8.3*) has significantly increased over 15 years, from 2018 onwards, trends of new opioid users appeared to stabilise/decrease (Figure 1). The year 2018 was found to be a significant decreasing change point in the trends of new opioid users for RA, AxSpA, and SLE, whilst this was 2013 in OA and 2019 for fibromyalgia. Opioid prescription rates among new opioid users increased in SLE (4.3# vs 5.4#), OA (4.6# vs 4.9#) and fibromyalgia (5.6# vs 6.5#) but decreased in RA (5.7# vs 5.3#) from 2006 to 2020, despite fluctuations in the rates observed in this period. The highest proportions of recurrent opioid users among the 6 RMDs were patients with RA (32.6%) and fibromyalgia (31.9%).

* The number of new opioid users per 10000 persons

# The number of opioid prescriptions in new users per patient years

Figure 1.

Trends of new opioid users by RMD, 2006-2020.

Conclusion

RA, PsA and fibromyalgia had an overall increase in new opioid users since 2006. The slight decrease in the trends of new opioid users among most of the RMDs after 2018 may reflect an increasing awareness of the opioid epidemic. The high proportions of recurrent opioid users in RA and fibromyalgia patients highlight the importance of exploring the safety of long-term opioid use and effective pain interventions for patients with RMDs.

References

[1]Jani M et al (2020) Time trends and prescribing patterns of opioid drugs in UK primary care patients with non-cancer pain: A retrospective cohort study. PLoS Med 17(10): e1003270

Acknowledgements

This work was funded by a FOREUM grant (grant ID: 125059), MJ is funded through an NIHR Advanced Fellowship (NIHR301413). Thanks to the CPRD fob holders in our centre, Ruth Costello and Ramiro Bravo, for downloading the data.

Disclosure of Interests

Joyce (Yun-Ting) Huang: None declared, David Jenkins: None declared, Belay Birlie Yimer: None declared, Jose Benitez-Aurioles: None declared, Niels Peek: None declared, Mark Lunt: None declared, William Dixon Consultant of: WGD has received consultancy fees from Abbvie and Google., Meghna Jani: None declared.

Therapeutic drug monitoring of biopharmaceuticals in inflammatory rheumatic and musculoskeletal disease: a systematic literature review informing EULAR points to consider

The objectives of this review were to collect and summarise evidence on therapeutic drug monitoring (TDM) of biopharmaceuticals in inflammatory rheumatic and musculoskeletal diseases and to inform the EULAR Task Force for the formulation of evidence-based points to consider. A systematic literature review (SLR) was performed, covering technical aspects and (clinical) utility of TDM, to answer 13 research questions. MEDLINE, Embase and Cochrane were searched until July 2020. American College of Rheumatology and EULAR abstracts were also considered for inclusion. Data were extracted in evidence tables and risk of bias assessment was performed. For the search on technical aspects, 678 records were identified, of which 22 papers were selected. For the clinical utility search, 3846 records were identified, of which 108 papers were included. Patient-related factors associated with biopharmaceutical blood concentrations included body weight, methotrexate comedication and disease activity. The identification of a target range was hampered by study variability, mainly disease activity measures and study type. Evidence was inconsistent for multiple clinical situations in which TDM is currently applied. However, for some particular scenarios, including prediction of future treatment response, non-response to treatment, tapering and hypersensitivity reactions, robust evidence was found. There is currently no evidence for routine use of proactive TDM, in part because published cost-effectiveness analyses do not incorporate the current landscape of biopharmaceutical costs and usage. This SLR yields evidence in favour of TDM of biopharmaceuticals in some clinical scenarios, but evidence is insufficient to support implementation of routine use of TDM.

2022 EULAR points to consider for remote care in rheumatic and musculoskeletal diseases

BACKGROUND

Remote care and telehealth have the potential to expand healthcare access, and the COVID-19 pandemic has called for alternative solutions to conventional face-to-face follow-up and monitoring. However, guidance is needed on the integration of telehealth into clinical care of people with rheumatic and musculoskeletal diseases (RMD).

OBJECTIVE

To develop EULAR points to consider (PtC) for the development, prioritisation and implementation of telehealth for people with RMD.

METHODS

A multidisciplinary EULAR task force (TF) of 30 members from 14 European countries was established, and the EULAR standardised operating procedures for development of PtC were followed. A systematic literature review was conducted to support the TF in formulating the PtC. The level of agreement among the TF was established by anonymous online voting.

RESULTS

Four overarching principles and nine PtC were formulated. The use of telehealth should be tailored to patient's needs and preferences. The healthcare team should have adequate equipment and training and have telecommunication skills. Telehealth can be used in screening for RMD as preassessment in the referral process, for disease monitoring and regulation of medication dosages and in some non-pharmacological interventions. People with RMD should be offered training in using telehealth, and barriers should be resolved whenever possible.The level of agreement to each statement ranged from 8.5 to 9.8/10.

CONCLUSION

The PtC have identified areas where telehealth could improve quality of care and increase healthcare access. Knowing about drivers and barriers of telehealth is a prerequisite to successfully establish remote care approaches in rheumatologic clinical practice.

P104 Comparative risk of infection for first-line csDMARD therapy in rheumatoid arthritis: a multinational cohort analysis of real world data

Background/Aims

Infections on conventional synthetic disease modifying anti-rheumatic drugs (csDMARDs) are an important concern for rheumatoid arthritis (RA) patients, especially during the COVID-19 pandemic. However comparative safety data between csDMARDs have been conflicting and limited in power. The objective was to assess the comparative safety of serious, opportunistic and all infections (including non-serious) of first-line csDMARDs in RA through a large multinational observational study.

Methods

We evaluated first-line new users of methotrexate (MTX), hydroxychloroquine (HCQ), sulfasalazine (SZ) and leflunomide (LEF) monotherapy. Data was obtained from four US databases (IQVIA US Ambulatory EMR (AMBER), Optum® De-identified Clinformatics® Datamart (Optum), IBM MarketScan® Medicare Supplemental Database (MDCR), and IBM MarketScan Commercial Database (CCAE)), one from Germany (IQVIA Disease Analyser Germany EMR (Germany)), and another from the UK (IQVIA UK The Health Improvement Network). Patients included were ≥18 years with a RA diagnosis between 2005-2019, without prior inflammatory arthritis, cancer or infection (in the preceding 30 days). Serious infections were defined as those requiring hospitalisation or resulting in death within 30 days; opportunistic infections were defined as per published EULAR consensus. Patients were followed from 1-day following treatment initiation to the earliest of treatment discontinuation, switching, or add-on plus 14 days, or loss to follow-up. Cox proportional-hazards models for MTX against each csDMARD with large-scale propensity score stratification were performed. A large set of negative control outcomes were used to calibrate hazard ratios (cHR) to account for potential residual confounding. Estimates were pooled where homogeneity across sources was adequate (I2&lt;0.4).

Results

A total of 247,511 patients were included (MTX: 141,647; HCQ: 73,286, SSZ: 16,521, LEF: 16,057), with pooled incidence rates of serious, opportunistic and all infections across sources for MTX users of 33.7, 20.1 and 311.8 per 1,000 pyrs, respectively. With MTX as the referent, for all infections, the pooled cHR (with 95% Confidence Intervals) for SSZ was 0.73 (0.62, 0.86); HCQ, 0.96 (0.89, 1.04); and LEF, 0.74 (0.50, 1.08). The serious infection pooled cHR for SSZ was 0.75 (0.58, 0.97) and for LEF, 0.93 (0.61, 1.40). For opportunistic infections, pooled cHR for HCQ was 1.04 (0.92, 1.19).

Conclusion

SSZ, LEF and less consistently HCQ had a lower risk of all (including non-serious) infections, compared to MTX. SSZ and LEF were associated with a 25% reduction in the expected risk of all infections. SSZ was associated with a 25% lower risk of serious infections relative to MTX. In the first large scale observational network study assessing comparative risk of infection with csDMARDs there were differences between drugs in risk for all infections, with potential implications for clinical care.

Disclosure

M. Jani: None. E. Burn: None. J. Weaver: Corporate appointments; Janssen. L. Carmona: None. K. Chatzidionysiou: None. B. Illigen: None. D. Vizcaya: Corporate appointments; Bayer. T. Duarte-Salles: None. P. Ryan: Corporate appointments; Janssen. D. Prieto-Alhambra: None.

Opioid prescribing among new users for non-cancer pain in the USA, Canada, UK, and Taiwan: A population-based cohort study

BACKGROUND

The opioid epidemic in North America has been driven by an increase in the use and potency of prescription opioids, with ensuing excessive opioid-related deaths. Internationally, there are lower rates of opioid-related mortality, possibly because of differences in prescribing and health system policies. Our aim was to compare opioid prescribing rates in patients without cancer, across 5 centers in 4 countries. In addition, we evaluated differences in the type, strength, and starting dose of medication and whether these characteristics changed over time.

METHODS AND FINDINGS

We conducted a retrospective multicenter cohort study of adults who are new users of opioids without prior cancer. Electronic health records and administrative health records from Boston (United States), Quebec and Alberta (Canada), United Kingdom, and Taiwan were used to identify patients between 2006 and 2015. Standard dosages in morphine milligram equivalents (MMEs) were calculated according to The Centers for Disease Control and Prevention. Age- and sex-standardized opioid prescribing rates were calculated for each jurisdiction. Of the 2,542,890 patients included, 44,690 were from Boston (US), 1,420,136 Alberta, 26,871 Quebec (Canada), 1,012,939 UK, and 38,254 Taiwan. The highest standardized opioid prescribing rates in 2014 were observed in Alberta at 66/1,000 persons compared to 52, 51, and 18/1,000 in the UK, US, and Quebec, respectively. The median MME/day (IQR) at initiation was highest in Boston at 38 (20 to 45); followed by Quebec, 27 (18 to 43); Alberta, 23 (9 to 38); UK, 12 (7 to 20); and Taiwan, 8 (4 to 11). Oxycodone was the first prescribed opioid in 65% of patients in the US cohort compared to 14% in Quebec, 4% in Alberta, 0.1% in the UK, and none in Taiwan. One of the limitations was that data were not available from all centers for the entirety of the 10-year period.

CONCLUSIONS

In this study, we observed substantial differences in opioid prescribing practices for non-cancer pain between jurisdictions. The preference to start patients on higher MME/day and more potent opioids in North America may be a contributing cause to the opioid epidemic.

Translating research into clinical practice: quality improvement to halve non-adherence to methotrexate

Objective

MTX remains the cornerstone for therapy for RA, yet research shows that non-adherence is significant and correlates with response to therapy. This study aimed to halve self-reported non-adherence to MTX at the Kellgren Centre for Rheumatology.

Methods

An anonymous self-report adherence questionnaire was developed and data collected for 3 months prior to the introduction of interventions, and then regularly for the subsequent 2.5 years. A series of interventions were implemented, including motivational interviewing training, consistent information about MTX and development of a summary bookmark. Information on clinic times was collected for consultations with and without motivational interviewing. Surveys were conducted to ascertain consistency of messages about MTX. A biochemical assay was used to test MTX serum levels in patients at two time points: before and 2.8 years following introduction of the changes. Remission rates at 6 and 12 months post-MTX initiation were retrieved from patient notes and cost savings estimated by comparing actual numbers of new biologic starters compared with expected numbers based on the numbers of consultants employed at the two time points.

Results

Between June and August 2016, self-reported non-adherence to MTX was 24.7%. Following introduction of the interventions, self-reported non-adherence rates reduced to an average of 7.4% between April 2018 and August 2019. Clinic times were not significantly increased when motivational interviewing was employed. Consistency of messages by staff across three key areas (benefits of MTX, alcohol guidance and importance of adherence) improved from 64% in September 2016 to 94% in January 2018. Biochemical non-adherence reduced from 56% (September 2016) to 17% (June 2019), whilst remission rates 6 months post-initiation of MTX improved from 13% in 2014/15 to 37% in 2017/18, resulting is estimated cost savings of £30 000 per year.

Conclusion

Non-adherence to MTX can be improved using simple measures including focussing on the adherence and the benefits of treatment, and providing consistent information across departments.

Glucocorticoid use is associated with an increased risk of hypertension

OBJECTIVES

Patients with RA are frequently treated with glucocorticoids (GCs), but evidence is conflicting about whether GCs are associated with hypertension. The aim of this study was to determine whether GCs are associated with incident hypertension in patients with RA.

METHODS

A retrospective cohort of patients with incident RA and without hypertension was identified from UK primary care electronic medical records (Clinical Practice Research Datalink). GC prescriptions were used to determine time-varying GC use, dose and cumulative dose, with a 3 month attribution window. Hypertension was identified through either: blood pressure measurements >140/90 mmHg, or antihypertensive prescriptions and a Read code for hypertension. Unadjusted and adjusted Cox proportional hazards regression models were fitted to determine whether there was an association between GC use and incident hypertension.

RESULTS

There were 17 760 patients in the cohort. A total of 7421 (42%) were prescribed GCs during follow-up. The incident rate of hypertension was 64.1 per 1000 person years (95% CI: 62.5, 65.7). The Cox proportional hazards model indicated that recent GC use was associated with a 17% increased hazard of hypertension (hazard ratio 1.17; 95% CI: 1.10, 1.24). When categorized by dose, only doses above 7.5 mg were significantly associated with hypertension. Cumulative dose did not indicate a clear pattern.

CONCLUSION

Recent GC use was associated with incident hypertension in patients with RA, in particular doses ≥7.5 mg were associated with hypertension. Clinicians need to consider cardiovascular risk when prescribing GCs, and ensure blood pressure is regularly monitored and treated where necessary.

Enzyme-linked immunosorbent assays for monitoring TNF-alpha inhibitors and antibody levels in people with rheumatoid arthritis: a systematic review and economic evaluation

BACKGROUND

Rheumatoid arthritis is a chronic autoimmune disease that primarily causes inflammation, pain and stiffness in the joints. People with severe disease may be treated with biological disease-modifying anti-rheumatic drugs, including tumour necrosis factor-α inhibitors, but the efficacy of these drugs is hampered by the presence of anti-drug antibodies. Monitoring the response to these treatments typically involves clinical assessment using response criteria, such as Disease Activity Score in 28 joints or European League Against Rheumatism. Enzyme-linked immunosorbent assays can also be used to measure drug and antibody levels in the blood. These tests may inform whether or not adjustments to treatment are required or help clinicians to understand the reasons for treatment non-response or a loss of response.

METHODS

Systematic reviews were conducted to identify studies reporting on the clinical effectiveness and cost-effectiveness of using enzyme-linked immunosorbent assays to measure drug and anti-drug antibody levels to monitor the response to tumour necrosis factor-α inhibitors [adalimumab (Humira^®; AbbVie, Inc., North Chicago, IL, USA), etanercept (Enbrel^®; Pfizer, Inc., New York, NY, USA), infliximab (Remicade^®, Merck Sharp & Dohme Limited, Hoddesdon, UK), certolizumab pegol (Cimzia^®; UCB Pharma Limited, Slough, UK) and golimumab (Simponi^®; Merck Sharp & Dohme Limited)] in people with rheumatoid arthritis who had either achieved treatment target (remission or low disease activity) or shown primary or secondary non-response to treatment. A range of bibliographic databases, including MEDLINE, EMBASE and CENTRAL (Cochrane Central Register of Controlled Trials), were searched from inception to November 2018. The risk of bias was assessed using the Cochrane ROBINS-1 (Risk Of Bias In Non-randomised Studies - of Interventions) tool for non-randomised studies, with adaptations as appropriate. Threshold and cost-utility analyses that were based on a decision tree model were conducted to estimate the economic outcomes of adding therapeutic drug monitoring to standard care. The costs and resource use were considered from the perspective of the NHS and Personal Social Services. No discounting was applied to the costs and effects owing to the short-term time horizon of 18 months that was adopted in the economic analysis. The impact on the results of variations in testing and treatment strategies was explored in numerous clinically plausible sensitivity analyses.

RESULTS

Two studies were identified: (1) a non-randomised controlled trial, INGEBIO, that compared standard care with therapeutic drug monitoring using Promonitor^® assays [Progenika Biopharma SA (a Grifols-Progenika company), Derio, Spain] in Spanish patients receiving adalimumab who had achieved remission or low disease activity; and (2) a historical control study. The economic analyses were informed by INGEBIO. Different outcomes from INGEBIO produced inconsistent results in both threshold and cost-utility analyses. The cost-effectiveness of therapeutic drug monitoring varied, from the intervention being dominant to the incremental cost-effectiveness ratio of £164,009 per quality-adjusted life-year gained. However, when the frequency of testing was assumed to be once per year and the cost of phlebotomy appointments was excluded, therapeutic drug monitoring dominated standard care.

LIMITATIONS

There is limited relevant research evidence and much uncertainty about the clinical effectiveness and cost-effectiveness of using enzyme-linked immunosorbent assay-based testing for therapeutic drug monitoring in rheumatoid arthritis patients. INGEBIO had serious limitations in relation to the National Institute for Health and Care Excellence scope: only one-third of participants had rheumatoid arthritis, the analyses were mostly not by intention to treat and the follow-up was 18 months only. Moreover, the outcomes might not be generalisable to the NHS.

CONCLUSIONS

Based on the available evidence, no firm conclusions could be made about the cost-effectiveness of therapeutic drug monitoring in England and Wales.

FUTURE WORK

Further controlled trials are required to assess the impact of using enzyme-linked immunosorbent assays for monitoring the anti-tumour necrosis factors in people with rheumatoid arthritis.

STUDY REGISTRATION

This study is registered as PROSPERO CRD42018105195.

FUNDING

This project was funded by the National Institute for Health Research (NIHR) Health Technology Assessment programme and will be published in full in Health Technology Assessment; Vol. 25, No. 8. See the NIHR Journals Library website for further project information.

Time trends and prescribing patterns of opioid drugs in UK primary care patients with non-cancer pain: A retrospective cohort study

BACKGROUND

The US opioid epidemic has led to similar concerns about prescribed opioids in the UK. In new users, initiation of or escalation to more potent and high dose opioids may contribute to long-term use. Additionally, physician prescribing behaviour has been described as a key driver of rising opioid prescriptions and long-term opioid use. No studies to our knowledge have investigated the extent to which regions, practices, and prescribers vary in opioid prescribing whilst accounting for case mix. This study sought to (i) describe prescribing trends between 2006 and 2017, (ii) evaluate the transition of opioid dose and potency in the first 2 years from initial prescription, (iii) quantify and identify risk factors for long-term opioid use, and (iv) quantify the variation of long-term use attributed to region, practice, and prescriber, accounting for case mix and chance variation.

METHODS AND FINDINGS

A retrospective cohort study using UK primary care electronic health records from the Clinical Practice Research Datalink was performed. Adult patients without cancer with a new prescription of an opioid were included; 1,968,742 new users of opioids were identified. Mean age was 51 ± 19 years, and 57% were female. Codeine was the most commonly prescribed opioid, with use increasing 5-fold from 2006 to 2017, reaching 2,456 prescriptions/10,000 people/year. Morphine, buprenorphine, and oxycodone prescribing rates continued to rise steadily throughout the study period. Of those who started on high dose (120-199 morphine milligram equivalents [MME]/day) or very high dose opioids (≥200 MME/day), 10.3% and 18.7% remained in the same MME/day category or higher at 2 years, respectively. Following opioid initiation, 14.6% became long-term opioid users in the first year. In the fully adjusted model, the following were associated with the highest adjusted odds ratios (aORs) for long-term use: older age (≥75 years, aOR 4.59, 95% CI 4.48-4.70, p < 0.001; 65-74 years, aOR 3.77, 95% CI 3.68-3.85, p < 0.001, compared to <35 years), social deprivation (Townsend score quintile 5/most deprived, aOR 1.56, 95% CI 1.52-1.59, p < 0.001, compared to quintile 1/least deprived), fibromyalgia (aOR 1.81, 95% CI 1.49-2.19, p < 0.001), substance abuse (aOR 1.72, 95% CI 1.65-1.79, p < 0.001), suicide/self-harm (aOR 1.56, 95% CI 1.52-1.61, p < 0.001), rheumatological conditions (aOR 1.53, 95% CI 1.48-1.58, p < 0.001), gabapentinoid use (aOR 2.52, 95% CI 2.43-2.61, p < 0.001), and MME/day at initiation (aOR 1.08, 95% CI 1.07-1.08, p < 0.001). After adjustment for case mix, 3 of the 10 UK regions (North West [16%], Yorkshire and the Humber [15%], and South West [15%]), 103 practices (25.6%), and 540 prescribers (3.5%) had a higher proportion of patients with long-term use compared to the population average. This study was limited to patients prescribed opioids in primary care and does not include opioids available over the counter or prescribed in hospitals or drug treatment centres.

CONCLUSIONS

Of patients commencing opioids on very high MME/day (≥200), a high proportion stayed in the same category for a subsequent 2 years. Age, deprivation, prescribing factors, comorbidities such as fibromyalgia, rheumatological conditions, recent major surgery, and history of substance abuse, alcohol abuse, and self-harm/suicide were associated with long-term opioid use. Despite adjustment for case mix, variation across regions and especially practices and prescribers in high-risk prescribing was observed. Our findings support greater calls for action for reduction in practice and prescriber variation by promoting safe practice in opioid prescribing.

Latent Class Trajectory Modeling of 2-Component Disease Activity Score in 28 Joints Identifies Multiple Rheumatoid Arthritis Phenotypes of Response to Biologic Disease-Modifying Antirheumatic Drugs

OBJECTIVE

To determine whether using a reweighted disease activity score that better reflects joint synovitis, i.e., the 2-component Disease Activity Score in 28 joints (DAS28) (based on swollen joint count and C-reactive protein level), produces more clinically relevant treatment outcome trajectories compared to the standard 4-component DAS28.

METHODS

Latent class mixed modeling of response to biologic treatment was applied to 2,991 rheumatoid arthritis (RA) patients in whom treatment with a biologic disease-modifying antirheumatic drug was being initiated within the Biologics in Rheumatoid Arthritis Genetics and Genomics Study Syndicate cohort, using both 4-component and 2-component DAS28 scores as outcome measures. Patient groups with similar trajectories were compared in terms of pretreatment baseline characteristics (including disability and comorbidities) and follow-up characteristics (including antidrug antibody events, adherence to treatments, and blood drug levels). We compared the trajectories obtained using the 4- and 2-component scores to determine which characteristics were better captured by each.

RESULTS

Using the 4-component DAS28, we identified 3 trajectory groups, which is consistent with previous findings. We showed that the 4-component DAS28 captures information relating to depression. Using the 2-component DAS28, 7 trajectory groups were identified; among them, distinct groups of nonresponders had a higher incidence of respiratory comorbidities and a higher proportion of antidrug antibody events. We also identified a group of patients for whom the 2-component DAS28 scores remained relatively low; this group included a high percentage of patients who were nonadherent to treatment. This highlights the utility of both the 4- and 2-component DAS28 for monitoring different components of disease activity.

CONCLUSION

Here we show that the 2-component modified DAS28 defines important biologic and clinical phenotypes associated with treatment outcome in RA and characterizes important underlying response mechanisms to biologic drugs.

OP0216 DEVELOPMENT AND VALIDATION OF PATIENT-LEVEL PREDICTION MODELS FOR ADVERSE HEALTH OUTCOMES AMONGST ADULT RA PATIENTS INITIATING FIRST-LINE TREATMENT OF METHOTREXATE MONOTHERAPY: A MULTINATIONAL REAL-WORLD COHORT ANALYSIS INCLUDING 164,735 SUBJECTS

Background:

EULAR guidelines recommend the early initiation of methotrexate (MTX) monotherapy as soon as possible after the diagnosis of rheumatoid arthritis (RA). Evaluating patient-level risks for adverse outcomes after MTX initiation would allow clinicians to provide more personalised care.

Objectives:

To develop and validate patient-level prediction models for adverse health outcomes including leukopenia, pancytopenia, infection (serious, opportunistic, all), cardiovascular disease (CVD) (myocardial infarction (MI), stroke), and cancer (breast, colorectal, uterus) in adult RA patients initiating first-line treatment of MTX monotherapy

Methods:

Health data from claims and electronic health records were used including patients from 7 European countries (Spain, Estonia, Netherlands, Belgium, Germany, France, and the UK), the United States of America, Australia, and Japan. All RA patients initiating first-line treatment of MTX monotherapy with at least one year of prior observation were included. Prediction models for the outcomes were developed for a time at risk of 3 months (infections, leukopenia, pancytopenia), 2 years (MI and stroke), and 5 years (cancers) on the Optum© De-Identified Clinformatics® Data Mart Database. Models were developed using LASSO logistic regression and were evaluated using the area under the receiver operator characteristic curve (AUROC) for discrimination and graphically assessed for calibration. The models were externally validated on all other databases.

Results:

A total of 21,307 subjects were used for training and validated against 143,427 patients from 14 sites. MI (AUROC internal 0.77, AUROC external ranging from 0.49 to 0.78), stroke (AUROC internal 0.78, AUROC external ranging from 0.68 to 0.79) and serious infection (AUROC internal 0.75, AUROC external ranging from 0.63 to 0.79) had good predictive validity [Table 1]. Discrimination for all other outcomes was lower, with all AUC<0.7 in internal validation. For detailed results see:https://data.ohdsi.org/ehdenRaPrediction/

Table 1.

Internal (Optum) and external validation results: AUC ROC for discrimination

DatabaseAcute MI within 2yStroke within 2ySerious Infection within 3mOptum (internal)0.770.780.75PanTher0.760.780.74IQVIA_AMBEMR0.760.72CCAE0.730.730.66IQVIA_GERMANY0.640.70IQVIA_THIN0.620.65MDCR0.680.680.67IQVIA_HOSPITAL0.670.630.61MDCD0.720.790.63JMDC0.490.750.71IQVIA_LPDFRANCE0.69Estonia0.670.770.82IQVIA_AUS0.58IPCI0.68SIDIAP0.650.75

Conclusion:

Clinical tools were developed that successfully identify subjects at risk of MI, stroke and serious infection at the initiation of first-line MTX therapy. The developed algorithms had good transportability and generally, the models with high AUROC had adequate internal calibration although some external validations show they could benefit from recalibration. For short-term opportunistic and all infections, as well as 5-year cancer models, we were unable to achieve a high enough AUROC to warrant validating externally.

Disclosure of Interests:

Cynthia Yang: None declared, Ross Williams: None declared, Joel Swerdel Shareholder of: J&J shares, Grant/research support from: Full-time employment salary from Janssen, Consultant of: Janssen employee, Employee of: Janssen, Paid instructor for: Janssen employee, have instructed at conferences, Speakers bureau: Janssen employee, have spoken at conferences, Meghna Jani Speakers bureau: Grifols, Talita Duarte-Salles: None declared, Katerina Chatzidionysiou Consultant of: AbbVie, Pfizer, Lilly., Daniel Prieto-Alhambra Grant/research support from: Professor Prieto-Alhambra has received research Grants from AMGEN, UCB Biopharma and Les Laboratoires Servier, Consultant of: DPA’s department has received fees for consultancy services from UCB Biopharma, Speakers bureau: DPA’s department has received fees for speaker and advisory board membership services from Amgen, Patrick Ryan: None declared, Peter Rijnbeek: None declared

FRI0120 ORAL GLUCOCORTICOID USE IS ASSOCIATED WITH HYPERTENSION IN PATIENTS WITH RHEUMATOID ARTHRITIS

Background:

Oral glucocorticoids (GC) are frequently prescribed to patients with rheumatoid arthritis (RA), however GC use is associated with a number of potential side effects. Hypertension is cited as a possible side effect, but few studies have specifically investigated GC-associated hypertension in patients with RA with conflicting results.

Objectives:

The aim of this study was to determine whether GCs were associated with an increased risk of incident hypertension in a cohort of patients with RA.

Methods:

A retrospective cohort of patients with incident RA and no hypertension at RA diagnosis were identified from UK primary care electronic health records (Clinical Practice Research Datalink). GC prescriptions were used to determine time-varying GC use and dose, categorised as: no use, >0–4.9 mg/day, 5–7.4 mg/day, 7.5–14.9 mg/day, ≥15mg/day. A 3-month risk attribution model was used where patients continued to remain at risk for 3 months after the end of prescriptions. Hypertension was identified if a patient had either: 1) 2 consecutive systolic blood pressure (BP) measurements >140mmHg within a year, 2) 2 consecutive diastolic BP measurements >90mmHg within a year or 3) antihypertensive prescriptions on at least two occasions and a Read code for hypertension. Unadjusted and adjusted Cox proportional hazards (PH) regression models were fitted to determine if there was an association between GC use and hypertension. Models were adjusted for baseline age, gender, baseline body mass index, baseline ever smoking, time-varying synthetic disease-modifying anti-rheumatic drug use, time-varying non-steroidal anti-inflammatory drug use and baseline Charlson comorbidity index.

Results:

There were 17,760 patients with incident RA and no hypertension. The cohort had a mean age of 56.3 ± 12.7 years and were predominantly female (68%). 7,421 (42%) were prescribed GCs during follow-up. There were 6,243 cases of incident hypertension over 97547 person years (pyrs) of follow-up, giving an incident rate of 64.1 per 1000 pyrs. Of those 1321 cases were in those exposed to GCs and 4922 were in those unexposed, giving incident rates of 87.6 per 1000 pyrs and 59.7 per 1000 pyrs, respectively. The adjusted Cox PH model indicated that recent GC use was associated with a 17% increased hazard of hypertension (hazard ratio: 1.17 (95% CI 1.10 to 1.24)). When categorised by dose, the adjusted model indicated only doses above 7.5mg were significantly associated with hypertension (Table 1).

Table 1.

Unadjusted and adjusted Cox proportional hazards regression model results

UnadjustedHR (95% CI)Age and gender adjustedHR (95% CI)Fully adjusted* HR (95% CI)Recent GC use1.44(1.35 to 1.53)1.23(1.16 to 1.31)1.17(1.10 to 1.24)Recent GC doseNo GC useReferenceReferenceReference>0 – 4.9mg1.35(1.21 to 1.53)1.13(1.01 to 1.28)1.10(0.98 to 1.24)5mg – 7.4mg1.40(1.22 to 1.60)1.11(0.97 to 1.27)1.07(0.93 to 1.23)7.5mg – 14.9mg1.44(1.33 to 1.57)1.26(1.16 to 1.38)1.18(1.08 to 1.29)15mg and over1.60(1.40 to 1.84)1.45(1.27 to 1.66)1.36(1.18 to 1.56)

* Adjusted for: Baseline age, gender, baseline body mass index, baseline ever smoking, synthetic disease-modifying anti-rheumatic drug use (time-varying), non-steroidal anti-inflammatory drug use (time-varying) and baseline Charlson comorbidity index.

Conclusion:

In this large cohort of patients with RA and without hypertension, recent GC use was associated with incident hypertension. In particular doses ≥7.5mg were associated with hypertension while the association with lower doses was inconclusive. Clinicians need to consider cardiovascular risk when prescribing GCs and ensure BP is regularly monitored.

Disclosure of Interests:

Ruth E Costello: None declared, Belay Birlie Yimer: None declared, Meghna Jani Speakers bureau: Grifols, William Dixon Consultant of: Bayer and Google

OP0087 NATIONAL VARIATION AND FACTORS ASSOCIATED WITH THE TRANSITION FROM FIRST USE TO LONG-TERM OPIOID USE FOR NON-CANCER PAIN

Background:

Prescribing behaviour of physicians has been described as a key driver of rising opioid prescriptions and long-term opioid use. However, the effect of prescribers requires interpretation within context. No studies have investigated the extent to which regions, practices and prescribers, vary in opioid prescribing accounting for case-mix by considering this hierarchy together.

Objectives:

(i) Quantify and identify risk factors for the transition from new-user to long-term opioid user (ii) Quantify variation of long-term use attributed to region, practice and prescriber, accounting for patient mix and chance variation.

Methods:

We conducted a retrospective observational UK study between 2006-2017 using Clinical Practice Research Datalink. Opioids new users, ≥18 years, without cancer were identified. Long-term opioid use was defined as ≥3 opioid prescriptions issued within a 90-day period from index date, or ≥1 opioid prescription lasting at least 90 days in the first year. A multi-level random-effects logistic regression model was used to examine the association of patient characteristics with the odds of becoming a long-term opioid user. To examine variation in opioid use amongst prescribers, GP practices and region after adjusting for patient case-mix, we used a nested random-effect structure. A ‘high-risk’ region, prescriber or practice was defined as those where the entire adjusted 95% CI lay above the population average.

Results:

1,968,742 new opioid users were included; 14.6% patients transitioned to long-term use. In the fully adjusted model, factors associated with higher odds of long-term opioid use included high morphine milligram equivalents (MME)/day at first prescription, older age, deprivation, fibromyalgia, rheumatological conditions, major surgery (Table). After adjustment for case-mix, the North-West, Yorkshire and South-West were found to be high-risk regions for long-term use. 103 practices (25.6%) and 540 prescribers (3.5%) were associated with a significantly higher risk of long-term use. The odds of becoming a long-term user for a patient belonging to these prescribers reached up to >3.5 times than the population average.

Conclusion:

Prescribing factors, age, deprivation and conditions including fibromyalgia and rheumatological conditions were associated with higher odds of long-term opioid use. In the first UK study evaluating long-term opioid prescribing with adjustment for patient-level characteristics, variation in regions and especially practices and prescribers were observed. Our findings support greater calls for action to reduce practice and prescriber variation by promoting safe practice in opioid prescribing.

Table.

Factors associated with long-term opioid use using a multi-level model accounting for clustering of individuals within prescriber, practice and region

Individual factorsAdjusted Odds Ratio (95% CI) *Prescribing factorsIndex daily MME >2007.59 (6.29, 9.16)Index daily MME 100-2001.12 (1.03, 1.21)Index daily MME 50-1001.58 (1.49, 1.68)Index daily MME <50RefGabapentinoid use2.51 (2.43, 2.60)Psychotropic use1.28 (1.17, 1.40)Age>754.35 (4.26, 4.45)65-753.57 (3.50, 3.65)55-653.03 (2.96, 3.09)35-551.91 (1.88, 1.95)Age <35RefDeprivation (Townsend score)Quintile 5 (Most deprived)1.54 (1.51, 1.57)Quintile 41.34 (1.31, 1.36)Quintile 31.20 (1.18, 1.22)Quintile 21.09 (1.07, 1.11)Quintile 1 (Least deprived)RefPre-existing conditions/ prior proceduresFibromyalgia1.81 (1.49, 2.20)Substance use disorder1.76 (1.70, 1.83)Suicide and self-harm1.56 (1.51, 1.61)Rheumatological conditions Ψ1.54 (1.49, 1.59)Alcohol abuse1.50 (1.45, 1.55)Depression1.28 (1.26, 1.30)Major Surgery1.09 (1.06, 1.13)

Abbreviations: MME, Morphine Milligram Equivalent; *p<0.05. Index daily MME/day is the MME/day at first prescription (MME= daily dose in milligrams X opioid conversion ratio). Ψ Defined by Charlson score including rheumatoid arthritis, SLE, myositis.

Disclosure of Interests:

Meghna Jani Speakers bureau: Grifols, Belay Birlie Yimer: None declared, Therese Sheppard: None declared, Mark Lunt: None declared, William Dixon Consultant of: Bayer and Google

O18 National variation and factors associated with long-term opioid use for non-cancer pain in the first year of use

Background

Physician prescribing behaviour has been described as a key driver of rising opioid prescriptions and long-term opioid use. However, the effect of prescribers requires interpretation within context. No studies have investigated the extent to which regions, practices, prescribers, vary in opioid prescribing by considering this hierarchy together, whilst accounting for case-mix. Objectives: (i)quantify and identify risk factors for the transition from new-users to long-term opioid users (ii) quantify variation of long-term use attributed to region, practice, prescriber, accounting for patient mix and chance variation.

Methods

We conducted a retrospective observational study between 2006-2017 using Clinical Practice Research Datalink. New users of opioids, ≥18 years, without cancer were identified. Long-term opioid use was defined as ≥ 3 opioid prescriptions within a 90-day period from index date, or ≥ 1 opioid prescription lasting at least 90-days in the first year. A multi-level random-effects logistic regression model was used to examine the association of patient characteristics with the odds of becoming a long-term opioid-user. To examine variation in opioid use among prescribers, GP-practices and region after adjusting for case-mix, we used a nested random-effect structure. A ‘high-risk’ region, prescriber or practice was defined as those where the entire adjusted 95% CI lay above population average.

Results

1,968,742 new opioid users were included; 14.6% transitioned to long-term use. In the fully-adjusted model, factors associated with higher-odds of long-term use included high morphine-milligram equivalents (MME)/day at first prescription, older-age, deprivation, fibromyalgia, rheumatological conditions and prior surgery (Table 1). After adjustment for case-mix, the North-West, Yorkshire and South-West were found to be high-risk regions for long-term use. 103 practices (25.6%) and 540 prescribers (3.5%) were associated with a significantly higher risk of long-term use. The odds of becoming a long-term user for patients belonging to these prescribers reached up to &gt; 3.5 times than the population average.

Conclusion

Prescribing factors, age, deprivation and conditions including fibromyalgia and rheumatological conditions were associated with higher odds of long-term opioid use. In the first UK study evaluating long-term opioid prescribing with patient-level characteristics adjustment, variation in regions, especially practices and prescribers were observed. Our findings support greater calls for action to reduce practice/prescriber variation by promoting safe practice in opioid-prescribing.

Disclosures

M. Jani: None. B. Birlie-Yimer: None. T. Sheppard: None. M. Lunt: None. W.G. Dixon: None.

P203 Oral glucocorticoid use is associated with hypertension in patients with RA

Background

Oral glucocorticoids (GC) are frequently prescribed to patients with rheumatoid arthritis (RA), however GC use is associated with several potential side effects. Hypertension is cited as a possible side effect, but few studies have specifically investigated GC-associated hypertension in patients with RA with conflicting results. The aim of this study was to determine whether GCs were associated with an increased risk of incident hypertension in a cohort of patients with RA.

Methods

A retrospective cohort of patients with incident RA and no hypertension at RA diagnosis were identified from UK primary care electronic health records (Clinical Practice Research Datalink). GC prescriptions were used to determine time-varying GC use and dose, categorised as: no use, &gt;0-4.9 mg/day, 5-7.4 mg/day, 7.5-14.9 mg/day, ≥15mg/day. A 3-month risk attribution model was used where patients continued to remain at risk for 3 months after the end of prescriptions. Hypertension was identified if a patient had either: 1) 2 consecutive systolic blood pressure (BP) measurements &gt;140mmHg within a year, 2) 2 consecutive diastolic BP measurements &gt;90mmHg within a year or 3) antihypertensive prescriptions on at least two occasions and a Read code for hypertension. Unadjusted and adjusted Cox proportional hazards (PH) regression models were fitted to determine if there was an association between GC use and hypertension.

Results

There were 17,760 patients with incident RA and no hypertension. The cohort had a mean age of 56.3 ± 12.7 years and were predominantly female (68%). 7,421 (42%) were prescribed GCs during follow-up. There were 6,243 cases of incident hypertension. The Cox PH model indicated that recent GC use was associated with a 17% increased hazard of hypertension (hazard ratio: 1.17 (95% CI 1.10 to 1.24)). When categorised by dose, only doses above 7.5mg were significantly associated with hypertension (Table 1).

Conclusion

In this large cohort of patients with RA and without hypertension, recent GC use was associated with incident hypertension. Doses ≥7.5mg were associated with hypertension while the association with lower doses was inconclusive. Clinicians need to consider cardiovascular risk when prescribing GCs and ensure BP is regularly monitored.

Disclosures

R.E. Costello None. M. Jani None. B.B. Yimer None. W.G. Dixon None.

The Genetics of Psoriasis and Psoriatic Arthritis

Psoriatic arthritis (PsA) is an inflammatory arthritis that manifests in 20-30% of patients diagnosed with psoriasis. Epidemiologic studies suggest a substantial genetic contribution to PsA. There is a strong need for genome-wide association studies on patients with PsA, including PsA-weighted or specific variants, and a need for a better understanding of the relevance of HLA alleles in disease expression. Interferon signaling and the nuclear factor-κB cascade are involved in PsA, and there are genetic differences between purely cutaneous psoriasis (PsC) and PsA. Psoriasis susceptibility genes for which putative functional coding variants in TYK2 and TRAF3IP2 are strongly associated with PsC and PsA, and neutrophil extracellular traps promote Th17 induction in an Act1 D10N-dependent fashion. Genomics and serological factors may also predict treatment response in tumor necrosis factor inhibitors (TNFi) in PsA, and genetics may play a role in treatment response to TNFi. Collaborations through the Group for Research and Assessment of Psoriasis and Psoriatic Arthritis (GRAPPA) are essential to increase study population size, which will enhance the ability to detect the genetic variants that create a predisposition to psoriatic disease and to predict response to biological therapy.

Association of Pharmacological Biomarkers with Treatment Response and Longterm Disability in Patients with Psoriatic Arthritis: Results from OUTPASS

Objective.

To identify (1) whether tumor necrosis factor inhibitor (TNFi) drug levels/anti-drug antibodies (ADAb) are associated with treatment response and disability in patients with psoriatic arthritis (PsA); and (2) the factors associated with TNFi drug levels.

Methods.

Patients were recruited from a national multicenter prospective cohort with longitudinal serum samples and 28-joint count Disease Activity Scores (DAS28)/Health Assessment Questionnaire (HAQ) measurement over 12 months.

Results.

Adalimumab (ADA) drug levels were significantly associated with ΔDAS28 (β 0.055, 95% CI 0.011–0.099; p = 0.014) and inversely with HAQ over 12 months (β −0.022, 95% CI −0.043 to −0.00063). Factors significantly associated with ADA drug levels were ADAb levels and body mass index.

Conclusion.

Drug level testing in ADA-initiated PsA patients may be useful in determining treatment response/disability over 12 months.

Drug safety and immunogenicity of tumour necrosis factor inhibitors: the story so far

TNF-α inhibitor (TNFi) therapies have transformed the treatment of several rheumatic musculoskeletal diseases. However, the majority of TNFi's are immunogenic and consequent anti-drug antibodies formation can impact on both treatment efficacy and safety. Several controversies exist in the area of immunogenicity of TNFis and drug safety. While anti-drug antibodies to TNFis have been described in association with infusion reactions; serious adverse events (AEs) such as thromboembolic events, lupus-like syndrome, paradoxical AEs, for example, vasculitis-like events and other autoimmune manifestations have also been reported. The expansion of the biologic armamentarium, new treatment strategies such as introduction/switching to biosimilars and cost-saving approaches such as TNFi tapering, may all have a potential impact on immunogenicity and clinical sequelae. In this review we evaluate how evolution of biologics relates to drug safety and immunogenicity, appraise relevant evidence from trials, spontaneous pharmacovigilance and observational studies and outline the areas of uncertainty that still exist.

Prediction of infection risk in rheumatoid arthritis patients treated with biologics: are we any closer to risk stratification?

PURPOSE OF REVIEW

There are currently several available biologics for rheumatoid arthritis (RA) with similar efficacy in most trials. A major consideration therefore in choosing a biologic, continues to be safety concerns such as infection. Considerable advances have been made in the understanding of biologic safety on a population level; however, how close are we to stratifying risk for individual patients? This review discusses evidence published in the last year, with reference to key previous literature.

RECENT FINDINGS

Comparative safety of biologics has been studied in observational cohorts, with a possible increased risk of serious infection in tocilizumab-treated patients compared with etanercept. Rheumatoid arthritis patients on biologics are often on concomitant medications such as steroids and opioids, and the advances in relation to infection are summarized. Pharmacological biomarkers and optimizing existing risk prediction scores may allow better future risk stratification.

SUMMARY

Improved quantification of personalized benefit:harms would allow better-informed decisions, reduction of infection-associated morbidity as well as direct/indirect costs associated with biologics. Although advances have been made to better understand and predict risk, future studies are likely to require a range of novel data sources and methodologies for the goal of precision medicine to be truly realized.

International Comparison in Opiate Prescribing for New Users in Primary Care using Electronic Medical Record Data

IntroductionThe opioid epidemic in North America has, in part, been attributed to an increase in opiate use for non-cancer pain and the prescription of more potent molecules. In contrast, the United Kingdom appears unaffected by this crisis, possibly because of differences in primary care prescribing, or health system policies. ObjectiveTo determine if there are differences in opiate prescribing for new users in primary care in the United Kingdom, United States, and Canada. ApproachElectronic health record data from Quebec, Canada (MOXXI), the United States (Partners Health Care, Boston MA), and the United Kingdom (CPRD random sample of 600,000) were used to identify new users of opiates (no prior prescription in 2 years), at least 18 years old between 2006-2016. Cancer patients were excluded after harmonizing equivalent READ and ICD9/10 codes. Generic drug names in each jurisdiction were mapped to the WHO ATC classification, and characterized using morphine milligram equivalents (MME). ResultsOverall 655,877 new users were identified, of whom 78% of 58,286 (U.S.), 88% of 6,251 (Canada), and 96% of 600,000 (UK) were non-cancer patients. Mean age of new users was 49 (SD 16) in the US, 57 (SD 16) in Canada, and 52 (SD 19) in the UK. 57.6% (UK) to 67.3% (US) of new users were women. In the UK, 86.5% of patients were started on codeine (MME:0.15), compared to 43.9% in Canada and 8.5% in the U.S. In the U.S 65.0\% were started on oxycodone (MME:1.5), and 10.9% on hydrocodone (MME:1). In Canada, tramadol (18.2%; MME: 0.1) followed by oxycodone (13.2%) were the next most commonly prescribed drugs. Conclusion/ImplicationsSubstantial differences in opioid prescribing practices for non-cancer pain were observed between the UK and Canadian and United States sites. The predilection to start patients on more potent opiates in North America may be a contributing cause to the opiate epidemic.

Frequent discussion of insomnia and weight gain with glucocorticoid therapy: An analysis of Twitter posts

In recent years, social media websites have been suggested as a novel, vast source of data which may be useful for deriving drug safety information. Despite this, there are few published reports of drug safety profiles derived in this way. The aims of this study were to detect and quantify glucocorticoid-related adverse events using a computerised system for automated detection of suspected adverse drug reactions (ADR) from narrative text in Twitter, and to compare the frequency of specific ADR mentions within Twitter to the frequency and patterns of spontaneous ADR reporting to a national drug regulatory body. Of 159,297 tweets mentioning either prednisolone or prednisone between 1st October 2012 and 30th June 2015, 20,206 tweets were deemed to contain information resembling an ADR. The top AE MedDRA® Preferred Terms were ‘insomnia’ and ‘weight increased’, both recognised non-serious but common side effects. These were proportionally over-reported in Twitter when compared to spontaneous reports in the UK regulator’s ADR reporting scheme. Serious glucocorticoid related AEs were reported less frequently. Pharmacovigilance using Twitter data has the potential to be a valuable, supplementary source of drug safety information. In particular, it can illustrate which drug side effects patients discuss most commonly, potentially because of important impacts on quality of life. This information could help clinicians to inform patients about frequent and relevant non-serious side effects as well as more serious side effects.

Patients on steroid drugs often complain of insomnia and weight gain on Twitter, offering a window into commonly experienced side effects. William Dixon from the University of Manchester, UK, and colleagues searched for mentions of either prednisolone or prednisone on Twitter between October 2012 and June 2015. They documented around 20,000 tweets on the social media platform that contained discussion of an adverse drug reaction. The top side effects cited were “insomnia” and “weight increased”, both of which constituted a larger proportion of total steroid-related complaints on Twitter than in the official UK regulator’s reporting scheme. Serious side effects were mentioned less frequently on Twitter. The findings show the power of social media data to identify the side effects impacting quality of life among patients.