Performing studies using the UK Clinical Practice Research Datalink: to link or not to link?

Long-term event rates, risk factors, and treatment pattern in 1.4 million individuals qualifying for dual blood pressure lowering therapy

OBJECTIVES

We assessed rates of cardiovascular events, all-cause death, baseline risk factors, and treatment patterns in a population qualifying for initiation of dual combination blood pressure (BP)-lowering therapy. We also evaluated the association between dual versus monotherapy during follow-up and incidence of cardiovascular events.

METHODS

This study utilized integrated databases in England: Clinical Practice Research Datalink, Hospital Episode Statistics, and Office for National Statistics. Individuals aged at least 18 years qualifying for dual therapy were identified during 15-year period (2005-2019). The primary endpoint was composite of nonfatal myocardial infarction, nonfatal stroke, hospitalization for heart failure, and cardiovascular death. The secondary endpoint was all-cause death.

RESULTS

Total 1 426 079 individuals met selection criteria. The 15-year event rates for the primary and secondary endpoints were 27.1 and 32.6%, respectively. Atherosclerotic cardiovascular disease, diabetes on insulin therapy, heart failure, atrial fibrillation, chronic kidney disease, and advanced age were associated with two to four-fold higher risk of primary and secondary endpoints. The estimated hazard ratio for dual versus monotherapy as a time-varying covariate was 0.82 (95% confidence interval 0.81-0.83) for the primary endpoint. At variance with guidelines, monotherapy was most common treatment pattern over 5-year follow-up.

CONCLUSION

Baseline characteristics conveying a multifold higher risk for cardiovascular events and all-cause death mostly represented nonmodifiable risk factors. Treatment with dual therapy as compared to monotherapy was associated with reduction in cardiovascular events. Monotherapy remained most common BP-lowering treatment indicating substantial opportunity for risk reduction by treatment intensification.

Cohort studies on 71 outcomes among people with atopic eczema in UK primary care data

Atopic eczema may be related to multiple subsequent adverse health outcomes. Here, we provide evidence to judge and compare associations between eczema and a comprehensive set of outcomes. We conducted 71 cohort studies (age, sex, general practice-matched) using Clinical Practice Research Datalink Aurum primary care records (1997-2023), comparing up to 3.6 million people with eczema to 16.8 million without. Eczema was associated with subsequent diagnosis of outcomes with adjusted hazard ratios (99% confidence intervals) from Cox regression of up to 4.02(3.95-4.10) for food allergy (rate difference [RD] per 1,000 person-years of 1.5). Besides strong associations with atopic and allergic conditions (e.g., asthma 1.87[1.39-1.82], RD5.4) and skin infections (e.g., molluscum contagiosum 1.81[1.64-1.96], RD1.8), the strongest associations were with Hodgkin's lymphoma (1.85[1.66-2.06], RD0.02), Alopecia Areata (1.77[1.71-1.83], RD0.2), Crohn's disease (1.62[1.54-1.69], RD0.1), Urticaria (1.58[1.57-1.60], RD1.9), Coeliac disease (1.42[1.37-1.47], RD0.1), Ulcerative colitis (1.40[1.34-1.46], RD0.1), Autoimmune liver disease (1.32[1.21-1.43], RD0.01), and Irritable bowel syndrome (1.31[1.29-1.32], RD0.7). Sensitivity analyses revealed the impact of consultation bias or choice of cohort age cut-off on findings. Comparatively large HRs in severe eczema were seen for some liver, gastrointestinal and cardiovascular conditions, osteoporosis, and fractures. Most cancers and neurological conditions were not associated with eczema.

R WE ready for reimbursement? A round up of developments in real-world evidence relating to health technology assessment: part 13

In this latest update we highlight a report from the European Medicines Agency on their use of real-world evidence (RWE) in decision making, RWE reporting guidance from the Canadian Agency for Drugs and Technologies in Health and highlight some new data demonstrating the value medicines for spinal muscular atrophy have brought patients.

Steroid prescribing in primary care increases prior to Hodgkin lymphoma diagnosis: A UK nationwide case-control study

BACKGROUND

Steroid use is associated with increased risk of Hodgkin lymphoma (HL). However, allergic symptoms commonly treated with steroids are also presenting features of HL in some patients, thereby introducing protopathic bias in estimates of aetiological associations. It is therefore important to examine steroid prescribing patterns pre-diagnosis to understand timing of associations and when healthcare use increases before cancer diagnosis to inform future epidemiological study design.

METHODS

We analysed steroid prescribing in 1232 HL patients and 7392 matched controls using primary care electronic health records (Clinical Practice Research Datalink (CPRD), 1987-2016). Using Poisson regression, we calculated monthly steroid prescribing rates for the 24-months preceding HL diagnosis, identifying the inflection point when they start to increase from baseline in cases, comparing rates with synchronous controls, and stratifying by route-of-administration and allergic disease status.

RESULTS

46 % of HL patients had a steroid prescription in the 24-months preceding diagnosis compared to 26 % of controls (OR 2.55, 95 %CI 2.25-2.89, p < 0.001). Odds of underlying HL were greatest in patients receiving multiple steroid prescriptions, oral steroids and in patients with a new allergic disease diagnosis. Among HL patients, steroid prescribing rates increased progressively from 7-months pre-diagnosis, doubling from 52 to 111 prescriptions/1000 patients/month.

CONCLUSION

Steroid prescribing increases during periods leading up to HL diagnosis, suggesting steroid-treated symptoms may be early presenting features of HL. A diagnostic window of appreciable length exists for potential earlier HL diagnosis in some patients; this 7-month 'lag-period' pre-diagnosis should be excluded in studies examining aetiological associations between steroids and HL.

Association between Inhaled β2-agonists Initiation and Risk of Major Adverse Cardiovascular Events: A Population-based Nested Case-Control Study

Purpose

Despite ample evidence underpinning the efficacy of β₂-agonists in asthma and chronic obstructive pulmonary disease (COPD), the occurrence of β₁- and β₂-adrenoceptors in the heart suggests that β₂-agonists may have deleterious cardiac effects. We investigated the association between new users of long-or short-acting β₂-agonists (LABA or SABA) or ICS (inhaled corticosteroids)/LABA and major adverse cardiovascular events (MACE).

Methods

A nested case–control analysis was conducted using the UK Clinical Practice Research Datalink of patients with asthma, COPD or asthma–COPD overlap with initial treatment of LABA, SABA, ICS/LABA, ICS, long-or short-acting muscarinic antagonist (LAMA or SAMA) between 01 January 1998 and 31 July 2018. The primary outcome was MACE, defined as the first occurrence of stroke, heart failure, myocardial infarction, arrhythmia, or cardiovascular death. Each case was matched with up to 10 controls on age, sex, date of cohort-entry, and duration of follow-up. The risk of MACE associated with β₂-agonists was estimated using conditional logistic regression after controlling for potential confounders.

Results

The cohort included 180,567 new users of β₂-agonists, ICS, SAMA, or LAMA. Among asthmatics, β₂-agonists were not associated with the risk of MACE (SABA vs ICS: HR 1.29 [0.96–1.73]; ICS/LABA vs ICS, HR 0.75 [0.33–1.73]). In contrast, among COPD patients, LABA (HR, 2.38 [1.04–5.47]), SABA (HR, 2.02 [1.13–3.59]) and ICS/LABA (HR, 2.08 [1.04–4.16]) users had an increased risk of MACE compared with SAMA users. Among patients with asthma–COPD overlap, SABA (HR, 2.57 [1.26–5.24]) was associated with an increased risk of MACE compared with ICS.

Conclusion

In conclusion, initiation of LABA, SABA, or ICS/LABA in COPD or SABA in asthma–COPD overlap is associated with increased risk of MACE. No associations were observed among patients with asthma.

Non-benzodiazepine hypnotic use for sleep disturbance in people aged over 55 years living with dementia: a series of cohort studies

BACKGROUND

Sleep disturbance affects around 60% of people living with dementia and can negatively affect their quality of life and that of their carers. Hypnotic Z-drugs (zolpidem, zopiclone and zaleplon) are commonly used to treat insomnia, but their safety and efficacy have not been evaluated for people living with dementia.

OBJECTIVES

To estimate the benefits and harms of Z-drugs in people living with dementia with sleep disturbance.

DESIGN

A series of observational cohort studies using existing data from (1) primary care linked to hospital admission data and (2) clinical cohort studies of people living with dementia.

DATA SOURCES

Primary care study - Clinical Practice Research Datalink linked to Hospital Episode Statistics and Office for National Statistics mortality data. Clinical cohort studies - the Resource Use and Disease Course in Dementia - Nursing Homes (REDIC) study, National Alzheimer's Coordinating Centre (NACC) clinical data set and the Improving Well-being and Health for People with Dementia (WHELD) in nursing homes randomised controlled trial.

SETTING

Primary care study - 371 primary care practices in England. Clinical cohort studies - 47 nursing homes in Norway, 34 Alzheimer's disease centres in the USA and 69 care homes in England.

PARTICIPANTS

Primary care study - NHS England primary care patients diagnosed with dementia and aged > 55 years, with sleep disturbance or prescribed Z-drugs or low-dose tricyclic antidepressants, followed over 2 years. Clinical cohort studies - people living with dementia consenting to participate, followed over 3 years, 12 years and 9 months, for REDIC, NACC and WHELD, respectively.

INTERVENTIONS

The primary exposure was prescription or use of Z-drugs. Secondary exposures included prescription or use of benzodiazepines, low-dose tricyclic antidepressants and antipsychotics.

MAIN OUTCOME MEASURES

Falls, fractures, infection, stroke, venous thromboembolism, mortality, cognitive function and quality of life. There were insufficient data to investigate sleep disturbance.

RESULTS

The primary care study and combined clinical cohort studies included 6809 and 18,659 people living with dementia, with 3089 and 914 taking Z-drugs, respectively. New Z-drug use was associated with a greater risk of fractures (hazard ratio 1.40, 95% confidence interval 1.01 to 1.94), with risk increasing with greater cumulative dose (p = 0.002). The hazard ratio for Z-drug use and hip fracture was 1.59 (95% confidence interval 1.00 to 2.53) and for mortality was 1.34 (95% confidence interval 1.10 to 1.64). No excess risks of falls, infections, stroke or venous thromboembolism were detected. Z-drug use also did not have an impact on cognition, neuropsychiatric symptoms, disability or quality of life.

LIMITATIONS

Primary care study - possible residual confounding because of difficulties in identifying patients with sleep disturbance and by dementia severity. Clinical cohort studies - the small numbers of people living with dementia taking Z-drugs and outcomes not necessarily being measured before Z-drug initiation restricted analyses.

CONCLUSIONS

We observed a dose-dependent increase in fracture risk, but no other harms, with Z-drug use in dementia. However, multiple outcomes were examined, increasing the risk of false-positive findings. The mortality association was unlikely to be causal. Further research is needed to confirm the increased fracture risk. Decisions to prescribe Z-drugs may need to consider the risk of fractures, balanced against the impact of improved sleep for people living with dementia and that of their carers. Our findings suggest that when Z-drugs are prescribed, falls prevention strategies may be needed, and that the prescription should be regularly reviewed.

FUTURE WORK

More research is needed on safe and effective management strategies for sleep disturbance in people living with dementia.

STUDY REGISTRATION

This study is registered as European Union electronic Register of Post-Authorisation Studies (EU PAS) 18006.

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. 1. See the NIHR Journals Library website for further project information.

Performance and clinical utility of supervised machine-learning approaches in detecting familial hypercholesterolaemia in primary care

Familial hypercholesterolaemia (FH) is a common inherited disorder, causing lifelong elevated low-density lipoprotein cholesterol (LDL-C). Most individuals with FH remain undiagnosed, precluding opportunities to prevent premature heart disease and death. Some machine-learning approaches improve detection of FH in electronic health records, though clinical impact is under-explored. We assessed performance of an array of machine-learning approaches for enhancing detection of FH, and their clinical utility, within a large primary care population. A retrospective cohort study was done using routine primary care clinical records of 4,027,775 individuals from the United Kingdom with total cholesterol measured from 1 January 1999 to 25 June 2019. Predictive accuracy of five common machine-learning algorithms (logistic regression, random forest, gradient boosting machines, neural networks and ensemble learning) were assessed for detecting FH. Predictive accuracy was assessed by area under the receiver operating curves (AUC) and expected vs observed calibration slope; with clinical utility assessed by expected case-review workload and likelihood ratios. There were 7928 incident diagnoses of FH. In addition to known clinical features of FH (raised total cholesterol or LDL-C and family history of premature coronary heart disease), machine-learning (ML) algorithms identified features such as raised triglycerides which reduced the likelihood of FH. Apart from logistic regression (AUC, 0.81), all four other ML approaches had similarly high predictive accuracy (AUC > 0.89). Calibration slope ranged from 0.997 for gradient boosting machines to 1.857 for logistic regression. Among those screened, high probability cases requiring clinical review varied from 0.73% using ensemble learning to 10.16% using deep learning, but with positive predictive values of 15.5% and 2.8% respectively. Ensemble learning exhibited a dominant positive likelihood ratio (45.5) compared to all other ML models (7.0-14.4). Machine-learning models show similar high accuracy in detecting FH, offering opportunities to increase diagnosis. However, the clinical case-finding workload required for yield of cases will differ substantially between models.

Comparative effectiveness of trastuzumab emtansine versus lapatinib plus chemotherapy for HER2+ metastatic breast cancer

Aim: To investigate the comparative effectiveness of trastuzumab emtansine (T-DM1) in a real-world population of HER2+ metastatic breast cancer (mBC) patients. Materials & methods: The Flatiron Health database was used to identify a cohort of HER2+ mBC patients who received first-line trastuzumab treatment and T-DM1 or lapatinib plus chemotherapy as second-line treatment. Overall survival was compared between the two groups. Results: A total of 278 patients with HER2+ mBC received second-line T-DM1 and 34 lapatinib plus chemotherapy. Overall survival was longer in patients treated with T-DM1 than those treated with lapatinib plus chemotherapy (adjusted hazard ratio: 0.56; 95% CI: 0.38-0.85). Conclusion: Real-world data supports the effectiveness of T-DM1 in the second-line treatment of HER2+ mBC patients.

Gender Differences in Inhaled Pharmacotherapy Utilization in Patients with Obstructive Airway Diseases (OADs): A Population-Based Study

Purpose

Gender differences in the incidence, susceptibility and severity of many obstructive airway diseases (OADs) have been well recognized. However, gender differences in the inhaled pharmacotherapy profile are not well characterized.

Methods

We conducted a retrospective cohort study to investigate gender differences in new-users of inhaled corticosteroids (ICS), short-or long-acting beta₂-agonist (SABA or LABA), ICS/LABA, short-or long-acting muscarinic antagonist (SAMA or LAMA) among patients with asthma, COPD or asthma-COPD overlap (ACO). We used Clinical Practice Research Datalink to identify OAD patients, 18 years and older, who were new-users (1-year washout period) from 01-January-1998 to 31-July-2018. Multivariable logistic regression was used to examine gender differences in each of the inhaled pharmacotherapies after controlling for potential confounders.

Results

A total of 242,079 new-users (asthma: 84.93%; COPD: 10.19%; ACO: 4.88%) of inhaled pharmacotherapies were identified. The multivariable analyses showed that males with COPD were more likely to be a new user of a LABA (odds ratio [OR] 1.29; 95% confidence interval [CI], 1.12–1.49), LAMA (OR 1.21; 95% CI 1.10–1.33), SAMA (OR 1.11; 95% CI 1.01–1.21) and less likely to be a new user of a SABA (OR 0.84; 95% CI, 0.80–0.89) compared to females. Similar patterns were also observed for patients with ACO; males were more likely to be prescribed with LABA (OR 1.26; 95% CI 1.03–1.55), LAMA (OR 1.28; 95% CI 1.11–1.48), SAMA (OR 1.28; 95% CI 1.11–1.48), and less likely to be a new user of a SABA (OR 0.89; 95% CI, 0.82–0.96). Also, males with asthma were more likely to be a new-user of ICS/LABA (OR 1.15; 95% CI, 1.08–1.23) and less likely to start an ICS (OR 0.97; 95% CI, 0.95–0.99) in comparison with females.

Conclusion

Our study showed significant gender differences in new-users of inhaled pharmacotherapies among OAD patients. Adjusting for proxies of disease severity, calendar year, smoking and socioeconomic status did not change the association by gender.

Objectives, design and main findings until 2020 from the Rotterdam Study

The Rotterdam Study is an ongoing prospective cohort study that started in 1990 in the city of Rotterdam, The Netherlands. The study aims to unravel etiology, preclinical course, natural history and potential targets for intervention for chronic diseases in mid-life and late-life. The study focuses on cardiovascular, endocrine, hepatic, neurological, ophthalmic, psychiatric, dermatological, otolaryngological, locomotor, and respiratory diseases. As of 2008, 14,926 subjects aged 45 years or over comprise the Rotterdam Study cohort. Since 2016, the cohort is being expanded by persons aged 40 years and over. The findings of the Rotterdam Study have been presented in over 1700 research articles and reports. This article provides an update on the rationale and design of the study. It also presents a summary of the major findings from the preceding 3 years and outlines developments for the coming period.

Prevalence of comorbid mental and physical illnesses and risks for self-harm and premature death among primary care patients diagnosed with fatigue syndromes

BACKGROUND

Fatigue syndromes (FSs) affect large numbers of individuals, yet evidence from epidemiological studies on adverse outcomes, such as premature death, is limited.

METHODS

Cohort study involving 385 general practices in England that contributed to the Clinical Practice Research Datalink (CPRD) with linked inpatient Hospital Episode Statistics (HES) and Office for National Statistics (ONS) cause of death information. A total of 10 477 patients aged 15 years and above, diagnosed with a FS during 2000-2014, were individually matched with up to 20 comparator patients without a history of having a FS. Prevalence ratios (PRs) were estimated to compare the FS and comparison cohorts on clinical characteristics. Adjusted hazard ratios (HRs) for subsequent adverse outcomes were estimated from stratified Cox regression models.

RESULTS

Among patients diagnosed with FSs, we found elevated baseline prevalence of: any psychiatric illness (PR 1.77; 95% CI 1.72-1.82), anxiety disorders (PR 1.92; 1.85-1.99), depression (PR 1.89; 1.83-1.96), psychotropic prescriptions (PR 1.68; 1.64-1.72) and comorbid physical illness (PR 1.28; 1.23-1.32). We found no significant differences in risks for: all-cause mortality (HR 0.99; 0.91-1.09), natural death (HR 0.99; 0.90-1.09), unnatural death (HR 1.00; 0.59-1.72) or suicide (HR 1.68; 0.78-3.63). We did, however, observe a significantly elevated non-fatal self-harm risk: HR 1.83; 1.56-2.15.

CONCLUSIONS

The absence of elevated premature mortality risk is reassuring. The raised prevalence of mental illness and increased non-fatal self-harm risk indicate a need for enhanced assessment and management of psychopathology associated with fatigue syndromes.

Predicting atrial fibrillation in primary care using machine learning

BACKGROUND

Atrial fibrillation (AF) is the most common sustained heart arrhythmia. However, as many cases are asymptomatic, a large proportion of patients remain undiagnosed until serious complications arise. Efficient, cost-effective detection of the undiagnosed may be supported by risk-prediction models relating patient factors to AF risk. However, there exists a need for an implementable risk model that is contemporaneous and informed by routinely collected patient data, reflecting the real-world pathology of AF.

METHODS

This study sought to develop and evaluate novel and conventional statistical and machine learning models for risk-predication of AF. This was a retrospective, cohort study of adults (aged ≥30 years) without a history of AF, listed on the Clinical Practice Research Datalink, from January 2006 to December 2016. Models evaluated included published risk models (Framingham, ARIC, CHARGE-AF), machine learning models, which evaluated baseline and time-updated information (neural network, LASSO, random forests, support vector machines), and Cox regression.

RESULTS

Analysis of 2,994,837 individuals (3.2% AF) identified time-varying neural networks as the optimal model achieving an AUROC of 0.827 vs. 0.725, with number needed to screen of 9 vs. 13 patients at 75% sensitivity, when compared with the best existing model CHARGE-AF. The optimal model confirmed known baseline risk factors (age, previous cardiovascular disease, antihypertensive medication usage) and identified additional time-varying predictors (proximity of cardiovascular events, body mass index (both levels and changes), pulse pressure, and the frequency of blood pressure measurements).

CONCLUSION

The optimal time-varying machine learning model exhibited greater predictive performance than existing AF risk models and reflected known and new patient risk factors for AF.

Comparing record linkage software programs and algorithms using real-world data

Linkage of medical databases, including insurer claims and electronic health records (EHRs), is increasingly common. However, few studies have investigated the behavior and output of linkage software. To determine how linkage quality is affected by different algorithms, blocking variables, methods for string matching and weight determination, and decision rules, we compared the performance of 4 nonproprietary linkage software packages linking patient identifiers from noninteroperable inpatient and outpatient EHRs. We linked datasets using first and last name, gender, and date of birth (DOB). We evaluated DOB and year of birth (YOB) as blocking variables and used exact and inexact matching methods. We compared the weights assigned to record pairs and evaluated how matching weights corresponded to a gold standard, medical record number. Deduplicated datasets contained 69,523 inpatient and 176,154 outpatient records, respectively. Linkage runs blocking on DOB produced weights ranging in number from 8 for exact matching to 64,273 for inexact matching. Linkage runs blocking on YOB produced 8 to 916,806 weights. Exact matching matched record pairs with identical test characteristics (sensitivity 90.48%, specificity 99.78%) for the highest ranked group, but algorithms differentially prioritized certain variables. Inexact matching behaved more variably, leading to dramatic differences in sensitivity (range 0.04–93.36%) and positive predictive value (PPV) (range 86.67–97.35%), even for the most highly ranked record pairs. Blocking on DOB led to higher PPV of highly ranked record pairs. An ensemble approach based on averaging scaled matching weights led to modestly improved accuracy. In summary, we found few differences in the rankings of record pairs with the highest matching weights across 4 linkage packages. Performance was more consistent for exact string matching than for inexact string matching. Most methods and software packages performed similarly when comparing matching accuracy with the gold standard. In some settings, an ensemble matching approach may outperform individual linkage algorithms.

Effectiveness of two-drug therapy versus monotherapy as initial regimen in hypertension: A propensity score-matched cohort study in the UK Clinical Practice Research Datalink

Purpose

To compare the effectiveness on blood pressure (BP) of initial two‐drug therapy versus monotherapy in hypertensive patients.

Methods

Using the Clinical Practice Research Datalink, linked with Hospital Episode Statistics and Office for National Statistics, we identified a cohort of adults with uncontrolled hypertension, initiating one or two antihypertensive drug classes between 2006 and 2014. New users of two drugs and monotherapy were matched 1:2 by propensity score. Main exposure was “as‐treated,” ie, until first regimen change. Primary and secondary endpoints were systolic and diastolic BP control and major adverse cardiovascular event (MACE), respectively. Hazard ratios (HR) and 95% confidence intervals (CI) were estimated using Cox proportional hazard models.

Results

Of 54 523 eligible patients, 3256 (6.0%) were initiated to a two‐drug combination. Of these, 2807 were matched to 5614 monotherapy users. Mean exposure duration was 12.7 months, with 76.5% patients changing their initial regimen. Two‐drug therapy was associated with a clinically significant BP control increase in all hypertensive patients (HR = 1.17 [95%CI: 1.09‐1.26]), more so in patients with grade 2‐3 hypertension (HR = 1.28 [1.17‐1.41]). An increase of 27% in BP control (HR = 1.27 [1.08‐1.49]) was observed in patients initiating an ACEi+CCB combination compared with initiators of either single class. No significant association was found between two‐drug therapy and MACE. Several sensitivity analyses confirmed the main findings.

Conclusions

Few patients initiated therapy with two drugs, reflecting UK guidelines' recommendation to start with monotherapy. This study supports the greater effectiveness of two‐drug therapy as the initial regimen for BP control.

The impact of smoking and smoking cessation on disease outcomes in ulcerative colitis: a nationwide population-based study

BACKGROUND

Smokers are less likely to develop ulcerative colitis (UC) but the impact of smoking and subsequent cessation on clinical outcomes in UC is unclear.

AIM

To evaluate the effect of smoking status and smoking cessation on disease outcomes.

METHODS

Using a nationally representative clinical research database, we identified incident cases of UC during 2005-2016. Patients were grouped as never-smokers, ex-smokers and smokers based on smoking status recorded in the 2 years preceding UC diagnosis. We defined subgroups of persistent smokers and smokers who quit within 2 years after diagnosis. We compared the rates of overall corticosteroid use, corticosteroid-requiring flares, corticosteroid dependency, thiopurine use, hospitalisation and colectomy between these groups.

RESULTS

We identified 6754 patients with a new diagnosis of UC over the study period with data on smoking status, of whom 878 were smokers at diagnosis. Smokers had a similar risk of corticosteroid-requiring flares (OR 1.16, 95% CI 0.92-1.25), thiopurine use (HR 0.84, 95% CI 0.62-1.14), corticosteroid dependency (HR 0.85, 95% CI 0.60-1.11), hospitalisation (HR 0.92, 95% CI 0.72-1.18) and colectomy (HR 0.78, 95% CI 0.50-1.21) in comparison with never-smokers. Rates of flares, thiopurine use, corticosteroid dependency, hospitalisation and colectomy were not significantly different between persistent smokers and those who quit smoking after a diagnosis of UC.

CONCLUSIONS

Smokers and never-smokers with UC have similar outcomes with respect to flares, thiopurine use, corticosteroid dependency, hospitalisation and colectomy. Smoking cessation was not associated with worse disease course. The risks associated with smoking outweigh any benefits. UC patients should be counselled against smoking.

Discontinuation and primary care visits in nonvalvular atrial fibrillation patients treated with apixaban or warfarin

Aim: Nonvalvular atrial fibrillation (NVAF) requires long-term anticoagulation treatment, which may necessitate frequent primary care visits. Materials & methods: NVAF patients initiating warfarin or apixaban in 2012–2017 were identified from linked primary (Clinical Practice Research Datalink) and secondary care (Hospital Episode Statistics) data. A propensity score matched Cox regression model compared discontinuation risk. Primary care visits were compared via negative binomial regression. Results: A total of 2695 apixaban users were matched to warfarin patients. Discontinuation risk was lower with apixaban than warfarin (hazard ratio: 0.40; 95% CI: 0.35–0.46). Apixaban patients averaged 12.2 annual primary care visits, versus 17.1 for warfarin users (p < 0.001). Conclusion: Apixaban was associated with reduced rates of discontinuation and primary care visits compared with warfarin.