Methods for time-varying exposure related problems in pharmacoepidemiology: An overview

Use of electronic health record data to examine administrations of pro re nata analgesics during hip fracture post-acute care

PURPOSE

Medications prescribed to older adults in US skilled nursing facilities (SNF) and administrations of pro re nata (PRN) "as needed" medications are unobservable in Medicare insurance claims. There is an ongoing deficit in our understanding of medication use during post-acute care. Using SNF electronic health record (EHR) datasets, including medication orders and barcode medication administration records, we described patterns of PRN analgesic prescribing and administrations among SNF residents with hip fracture.

METHODS

Eligible participants resided in SNFs owned by 11 chains, had a diagnosis of hip fracture between January 1, 2018 to August 2, 2021, and received at least one administration of an analgesic medication in the 100 days after the hip fracture. We described the scheduling of analgesics, the proportion of available PRN doses administered, and the proportion of days with at least one PRN analgesic administration.

RESULTS

Among 24 038 residents, 57.3% had orders for PRN acetaminophen, 67.4% PRN opioids, 4.2% PRN non-steroidal anti-inflammatory drugs, and 18.6% PRN combination products. The median proportion of available PRN doses administered per drug was 3%-50% and the median proportion of days where one or more doses of an ordered PRN analgesic was administered was 25%-75%. Results differed by analgesic class and the number of administrations ordered per day.

CONCLUSIONS

EHRs can be leveraged to ascertain precise analgesic exposures during SNF stays. Future pharmacoepidemiology studies should consider linking SNF EHRs to insurance claims to construct a longitudinal history of medication use and healthcare utilization prior to and during episodes of SNF care.

Cardiovascular safety of using non-steroidal anti-inflammatory drugs for gout: a Danish nationwide case-crossover study

Gout attacks are treated with uric-lowering and anti-inflammatory drugs. In patients with gout, non-steroidal anti-inflammatory drugs (NSAIDs) could be both cardiovascular beneficial, due to their anti-inflammatory actions, and cardiovascular hazardous, due to their prothrombotic, hypertensive, and proarrhythmic side effects. We, therefore, examined the risk of cardiovascular events associated with NSAID use in patients with gout. We conducted a nationwide, population-based case-crossover study of all Danes ≥ 18 years of age with first-time gout during 1997-2020, who experienced a cardiovascular event (myocardial infarction, ischemic stroke, congestive heart failure, atrial fibrillation/flutter, or cardiovascular death) (n = 59,150). The exposure was use of NSAIDs, overall and according to type (ibuprofen, naproxen, or diclofenac). We used the dates 300, 240, 180, and 120 before the outcome date as reference dates. We used the Mantel-Haenszel method to calculate odds ratios (ORs) with 95% confidence intervals (CIs) of the association between NSAID use and cardiovascular events. NSAID use was overall associated with 12% decreased odds of a cardiovascular event (OR = 0.88, 95% CI: 0.85-0.91). This decreased odds ratio was observed for the use of ibuprofen (OR = 0.92, 95% CI: 0.88-0.97) and naproxen (OR = 0.85, 95% CI: 0.74-0.97), but not for the use of diclofenac (OR = 0.97, 95% CI: 0.90-1.05). Overall, use of NSAIDs was associated with decreased odds of all the individual components of the composite outcome. NSAIDs were not associated with an increased cardiovascular event rate when used in gout patients. Ibuprofen and naproxen appeared to have better cardiovascular risk profiles than diclofenac.

Are patients with primary glomerular disease at increased risk of malignancy?

Over the past decade, several observational studies and case series have provided evidence suggesting a connection between glomerular diseases and the development of malignancies, with an estimated risk ranging from 5 to 11%. These malignancies include solid organ tumours as well as haematologic malignancies such as lymphoma and leukaemia. However, these risk estimates are subject to several sources of bias, including unmeasured confounding from inadequate exploration of risk factors, inclusion of glomerular disease cases that were potentially secondary to an underlying malignancy, misclassification of glomerular disease type and ascertainment bias arising from an increased likelihood of physician encounters compared with the general population. Consequently, population-based studies that accurately evaluate the cancer risk in glomerular disease populations are lacking. While it is speculated that long-term use of immunosuppressive medications and glomerular disease activity measured by proteinuria and estimated glomerular filtration rate may be associated with cancer risk in patients with glomerular disease, the independent role of these risk factors remains largely unknown. The presence of these knowledge gaps could lead to a lack of awareness of cancer as a potential chronic complication of glomerular disease, underutilization of routine screening practices in clinical care that allow early diagnosis and treatment of malignancies and underrecognition of modifiable risk factors to decrease the risk of de novo malignancies over time. This review summarizes the current evidence on the risk of cancer in patients with glomerular diseases, explores the limitations of prior studies and discusses methodological challenges and potential solutions for obtaining accurate estimates of cancer risk and identifying modifiable risk factors unique to GN populations.

Clinical outcomes following discontinuation of metformin in patients with type 2 diabetes and advanced chronic kidney disease in Hong Kong: a territory-wide, retrospective cohort and target trial emulation study

Background

Current labelling advises discontinuation of metformin when estimated glomerular filtration rate (eGFR) < 30 ml/min/1.73 m2 due to increased risk of lactic acidosis. However, in real-world practice, the risk-benefit ratios remain uncertain. We examined the risk associations of discontinued-metformin use with cardiorenal and clinical outcomes in patients with type 2 diabetes (T2D) and advanced chronic kidney disease.

Methods

In this territory-wide, retrospective cohort and target trial emulation study, we included Chinese patients attending the Hong Kong Hospital Authority (HA) and enrolled in the Risk-Assessment-and-Management-Programme-for-Diabetes-Mellitus (RAMP-DM) from 2002 to 2019. Patients were stratified by discontinuation of metformin within six months after reaching eGFR < 30 ml/min/1.73 m2 from January 1, 2002 to December 31, 2018, and followed up until December 31 2019. We excluded patients who had observational time <6 months from eGFR < 30 ml/min/1.73 m2, and had their eGFR measured during a hospitalisation episode due to acute kidney injury, or missing diagnosis date of diabetes. We compared the risk associations of metformin discontinuation with clinical outcomes. The primary outcomes were major adverse cardiovascular events (MACE), end-stage kidney disease (ESKD), cancer, and all-cause mortality. A Cox-model with time-dependent exposure and covariates was used to estimate the hazard ratio (HR) of outcomes in a propensity-score overlap-weighted cohort. The risk of occurrence of lactic acidosis (serum lactate > 5.0 mmol/L with a concomitant blood pH < 7.35 or ICD-9 codes of 276.2) in discontinued-metformin versus continued-metformin users was assessed in a separate register-based cohort.

Findings

A total of 33,586 metformin users with new-onset eGFR < 30 ml/min/1.73 m2 were included in the study, 7500 (22.3%) of whom discontinued metformin within 6 months whereas 26,086 (77.7%) continued use of metformin. During a median follow-up of 3.8 (IQR: 2.2-6.1) years, 16.4% (5505/33,586), 30.1% (10,113/33,586), and 7.1% (2171/30,682) had incident MACE, ESKD, and cancer respectively, and 44.4% (14,917/33,586) died. Compared to continued-metformin use, discontinuation was associated with higher risk of MACE (weighted and adjusted HR = 1.40, 95% CI: 1.29-1.52), ESKD (HR = 1.52, 1.42-1.62), and death (HR = 1.22, 1.18-1.27). No association was observed for cancer (HR = 0.93, 0.85-1.01). Discontinued-metformin users had higher change in HbA1c change at 6-month of follow-up versus continued-metformin users (weighted mean HbA1c level change: 0.5% [0.4-0.6%] versus 0.2% [0.1-0.2]). In the separate register-based cohort (n = 3235), null association was observed between metformin use and risk of lactic acidosis (weighted HR = 0.94 [0.53-1.64]).

Interpretation

Our results suggest that discontinuation of metformin in patients with T2D and chronic kidney disease may be associated with increased risk of cardiovascular-renal events. Use of metformin below eGFR of 30 ml/min/1.73 m2 may be associated with cardiovascular, renal, and mortality benefits that need to be weighed against the risk of lactic acidosis, but further research is needed to validate these findings.

Funding

CUHK Impact Research Fellowship Scheme.

Association between statin use and the risk for idiopathic pulmonary fibrosis and its prognosis: a nationwide, population-based study

Given the pleiotropic effects of statins beyond their lipid-lowering effects, there have been attempts to evaluate the role of statin therapy in IPF, but they have shown inconclusive results. Data from the National Health Insurance Service (NHIS) database of South Korea were used to investigate the effects of statin therapy on IPF. The IPF cohort consisted of a total of 10,568 patients who were newly diagnosed with IPF between 2010 and 2017. These patients were then matched in a 1:3 ratio to 31,704 subjects from a control cohort without IPF, with matching based on age and sex. A case-control study was performed to evaluate the association between statin use and the risk for IPF, and the multivariable analysis revealed that statin use was associated with a lower risk for IPF (adjusted OR 0.847, 95% CI 0.800-0.898). Using the IPF cohort, we also evaluated whether statin use at the time of diagnosis was associated with future clinical outcomes. The statin use at the time of IPF diagnosis was associated with improved overall survival (adjusted HR 0.779, 95% CI 0.709-0.856). Further prospective studies are needed to clarify the role of statin therapy in IPF.

Handling time-varying treatments in observational studies: A scoping review and recommendations

OBJECTIVE

Time-varying treatments are common in observational studies. However, when assessing treatment effects, the methodological framework has not been systematically established for handling time-varying treatments. This study aimed to examine the current methods for dealing with time-varying treatments in observational studies and developed practical recommendations.

METHODS

We searched PubMed from 2000 to 2021 for methodological articles about time-varying treatments, and qualitatively summarized the current methods for handling time-varying treatments. Subsequently, we developed practical recommendations through interactive internal group discussions and consensus by a panel of external experts.

RESULTS

Of the 36 eligible reports (22 methodological reviews, 10 original studies, 2 tutorials and 2 commentaries), most examined statistical methods for time-varying treatments, and only a few discussed the overarching methodological process. Generally, there were three methodological components to handle time-varying treatments. These included the specification of treatment which may be categorized as three scenarios (i.e., time-independent treatment, static treatment regime, or dynamic treatment regime); definition of treatment status which could involve three approaches (i.e., intention-to-treat, per-protocol, or as-treated approach); and selection of analytic methods. Based on the review results, a methodological workflow and a set of practical recommendations were proposed through two consensus meetings.

CONCLUSIONS

There is no consensus process for assessing treatment effects in observational studies with time-varying treatments. Previous efforts were dedicated to developing statistical methods. Our study proposed a stepwise workflow with practical recommendations to assist the practice.

Core concepts in pharmacoepidemiology: Measurement of medication exposure in routinely collected healthcare data for causal inference studies in pharmacoepidemiology

BACKGROUND

Observational designs can complement evidence from randomized controlled trials not only in situations when randomization is not feasible, but also by evaluating drug effects in real-world, considering a broader spectrum of users and clinical scenarios. However, use of such real-world scenarios captured in routinely collected clinical or administrative data also comes with specific challenges. Unlike in trials, medication use is not protocol based. Instead, exposure is determined by a multitude of factors involving patients, providers, healthcare access, and other policies. Accurate measurement of medication exposure relies on a similar broad set of factors which, if not understood and appropriately addressed, can lead to exposure misclassification and bias.

AIM

To describe core considerations for measurement of medication exposure in routinely collected healthcare data.

METHODS

We describe the strengths and weaknesses of the two main types of routinely collected healthcare data (electronic health records and administrative claims) used in pharmacoepidemiologic research. We introduce key elements in those data sources and issues in the curation process that should be considered when developing exposure definitions. We present challenges in exposure measurement such as the appropriate determination of exposure time windows or the delineation of concomitant medication use versus switching of therapy, and related implications for bias.

RESULTS

We note that true exposure patterns are typically unknown when using routinely collected healthcare data and that an in-depth understanding of healthcare delivery, patient and provider decision-making, data documentation and governance, as well as pharmacology are needed to ensure unbiased approaches to measuring exposure.

CONCLUSIONS

Various assumptions are made with the goal that the chosen exposure definition can approximate true exposure. However, the possibility of exposure misclassification remains, and sensitivity analyses that can test the impact of such assumptions on the robustness of estimated medication effects are necessary to support causal inferences.

Polypharmacy and risk of fractures in older adults: A systematic review

BACKGROUND

Fractures have serious health consequences in older adults. While some medications are individually associated with increased risk of falls and fractures, it is not clear if this holds true for the use of many medications (polypharmacy). We aimed to identify what is known about the association between polypharmacy and the risk of fractures in adults aged ≥65 and to examine the methods used to study this association.

METHODS

We conducted a systematic review with narrative synthesis of studies published up to October 2023 in PubMed, Embase, CINAHL, PsychINFO, Cochrane Library, Web of Science, and the grey literature. Two independent reviewers screened titles, abstracts, and full texts, then performed data extraction and quality assessment.

RESULTS

Among the 31 studies included, 11 different definitions of polypharmacy were used and were based on three medication counting methods (concurrent use 15/31, cumulative use over a period 6/31, daily average 3/31, and indeterminate 7/31). Overall, polypharmacy was frequent and associated with higher fracture risk. A dose-response relationship between increasing number of medications and increased risk of fractures was observed. However, only seven studies adjusted for major confounders (age, sex, and chronic disease). The quality of the studies ranged from poor to high.

CONCLUSIONS

Polypharmacy appears to be a relevant modifiable risk factor for fractures in older individuals that can easily be used to identify those at risk. The diversity of medication calculation methods and definitions of polypharmacy highlights the importance of a detailed methodology to understand and compare results.

Outcomes of Patients with COPD Treated with ICS/LABA Before and After Initiation of Single-Inhaler Triple Therapy with Fluticasone Furoate/Umeclidinium/Vilanterol (FF/UMEC/VI)

INTRODUCTION

Triple therapy (fluticasone furoate/umeclidinium/vilanterol; FF/UMEC/VI) has been shown to improve symptoms and reduce exacerbations in patients with chronic obstructive pulmonary disease (COPD) and a history of exacerbations. This real-world study compared exacerbation rates and healthcare resource utilization (HCRU) before and after initiation of FF/UMEC/VI in patients with COPD previously treated with inhaled corticosteroid (ICS)/long-acting β2-agonist (LABA).

METHODS

This retrospective cohort study included commercial and Medicare Advantage with Part D administrative claims data from September 01, 2016, to March 31, 2020, of patients diagnosed with COPD. The index date was the date of the first FF/UMEC/VI claim (September 2017-March 2019). The 12 months prior to index (baseline) were used to assess patient characteristics and outcomes; the 12 months following index (follow-up) were used to assess study outcomes. All patients had ≥ 30 consecutive days' supply of any ICS/LABA dual therapy during the 12 months prior to FF/UMEC/VI initiation. Subgroup analyses included patients with ≥ 30 consecutive days' supply of budesonide/formoterol (BUD/FORM) during baseline. Analyses of patients with ≥ 1 COPD exacerbation during baseline were reported as well.

RESULTS

The overall population included 1449 patients (mean age 70.75 years; 54.18% female), of whom 540 were patients in the BUD/FORM subgroup. Significantly fewer patients experienced any exacerbation during follow-up versus baseline (overall population 53.49% vs 62.59%; p < 0.001; BUD/FORM subgroup 55.00% vs 62.41%; p = 0.004). Effects on exacerbation reduction were more pronounced among patients with ≥ 1 exacerbation during baseline. Lower COPD-related HCRU was observed during the follow-up compared with baseline for both the overall population and the BUD/FORM subgroup.

CONCLUSION

Patients with COPD treated with ICS/LABA during baseline, including patients specifically treated with BUD/FORM and those with a history of ≥ 1 exacerbation, had fewer COPD exacerbations and lower COPD-related HCRU after initiating FF/UMEC/VI.

Changes in Glomerular Filtration Rate After Switching From Tenofovir Disoproxil Fumarate to Tenofovir Alafenamide Fumarate for Human Immunodeficiency Virus Preexposure Prophylaxis

Background

Tenofovir alafenamide fumarate (TAF) was promoted as a safer alternative to tenofovir disoproxil fumarate (TDF) for human immunodeficiency virus oral preexposure prophylaxis (PrEP). It is unknown if switching from TDF to TAF translates to improved renal function. We used electronic health record (EHR) data to assess changes in creatinine-estimated glomerular filtration rate (eGFR) after switching from TDF to TAF.

Methods

We conducted a retrospective cohort study using EHR data from Kaiser Permanente Southern California. We identified individuals who switched from TDF to TAF between October 2019 and May 2022 and used time-varying propensity score matching to identify controls who were on TDF ("nonswitchers"). We then used Bayesian longitudinal modeling to compare differences in eGFR between switching and nonswitching scenarios.

Results

Among 5246 eligible individuals, we included 118 TDF to TAF switchers and 114 nonswitchers. Compared to nonswitchers, switchers had older age of starting TDF but similar body weights at index date. A higher proportion of switchers were White, on Medicare or Medicaid, and had dyslipidemia at index date. Switching to TAF was associated with a higher eGFR compared to staying on TDF in 3-15 months post-switch, but the differences were not statistically significant (eg, month 9 difference: 1.27 [95% credible interval, -1.35 to 3.89]). While most of the estimated changes showed eGFR increase associated with switching, most were <2 eGFR units. Sensitivity analyses to address missingness or nonadherence showed similar results.

Conclusions

Switching from TDF to TAF for PrEP was associated with a nonsignificant increase in eGFR. Findings need to be confirmed using larger cohorts.

Calcium channel blockers and clinical outcomes in patients with continuous-flow left ventricular assist devices

AIMS

Current guidelines suggest calcium channel blockers (CCBs) as the second or third option for blood pressure management in patients with left ventricular assist device (LVAD). However, the clinical outcomes of patients with LVAD who receive CCBs remain unclear. Our study aims to analyse the association of CCBs with clinical outcomes in patients after LVAD implantation.

METHODS AND RESULTS

This is a retrospective analysis based on the Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS) from 2006 to 2017, and adult patients who were alive with LVAD and CCB treatment information at 6 months after implantation were included. Among 10 717 patients, 1369 received CCBs 6 months after implantation, and there was an increasing trend of CCB use after LVAD. Patients receiving CCB therapy at 6 months had a similar 5 year survival rate to those not receiving CCB [49.6%, 95% confidence interval (CI): 47.5-51.7% vs. 51.1%, 95% CI: 45.3-56.7%]. In both Cox and competing risk regressions after adjusting for confounding factors, CCB treatment at 6 months after implantation was not associated with long-term mortality [hazard ratio (HR): 1.03, 95% CI: 0.91-1.17, P = 0.624 and subdistribution HR (SHR): 1.07, 95% CI: 0.95-1.22, P = 0.260]. Consistently, in time-varying models, CCB treatment was not linked to long-term mortality (HR: 0.97, 95% CI: 0.87-1.09, P = 0.682 and SHR: 1.05, 95% CI: 0.94-1.18, P = 0.359). This null association remained in subgroup analysis according to device strategy and propensity-matching analyses. Neurological dysfunction, stroke, bleeding, rehospitalization, and renal dysfunction were more likely to occur among those with CCB when compared with those without CCB treatment.

CONCLUSIONS

In patients with LVAD, CCB therapy fails to show benefits in long-term survival and is associated with increased incidences of neurological dysfunction, bleeding, renal dysfunction, and rehospitalization.

The weighted cumulative exposure method and its application to pharmacoepidemiology: A narrative review

PURPOSE

The weighted cumulative exposure (WCE) method has been used in a number of fields including pharmacoepidemiology where it can account for intensity, duration and timing of exposures on the risk of an outcome. The method uses a data driven approach with flexible cubic B-splines to assign weights to past doses and select an aetiologically appropriate time window. Predictions of risk are possible for common exposure patterns encountered in real-world studies. The purpose of this study was to describe applications of the WCE method to pharmacoepidemiology and assess the strengths and limitations of the method.

METHOD

A literature search was undertaken to find studies applying the WCE method to the study of medicines. Articles published in PubMed using the search term 'weighted cumulative exposure' and articles citing Sylvestre et al. (2009) in Google Scholar or Scopus up to March 2023 were subsequently reviewed. Articles were selected based on title and review of abstracts.

RESULTS

Seventeen clinical applications using the data-driven WCE method with flexible cubic splines were identified in the review. These included 3 case-control studies and 14 cohort studies, of which 12 were analysed with Cox proportional hazards models and 2 with logistic regression. Thirteen studies used time windows of 1 year or longer. Of 11 studies which compared conventional models with the WCE method, 10 (91%) studies found a better fit with WCE models while one had an equivalent fit. The freely available 'WCE' software package has facilitated the applications of the WCE method with flexible cubic splines.

CONCLUSIONS

The WCE method allows additional insights into the effect of cumulative exposure on outcomes, including the timing and intensity (dose) of the exposure on the risk. The flexibility of the method is particularly well suited to studies with long-term exposures that vary over time or where the current risk of an event is affected by how far the exposure is in the past, which is difficult to model with conventional definitions of exposure. Interpretation of the results can be more complex than for conventional models and would be facilitated by a standardised reporting framework.

Non-Adherence to COPD Medications and its Association with Adverse Events: A Longitudinal Population Based Cohort Study of Older Adults

OBJECTIVE

To determine the association between non-adherence to long term chronic obstructive pulmonary disease (COPD) medications and COPD related emergency department (ED) visits and hospitalizations in patients with incident COPD, utilizing time varying measures of adherence as well as accounting for time-varying confounding impacted by prior adherence.

STUDY DESIGN AND SETTING

We conducted a population-based retrospective cohort study between 2007-2017 among individuals aged 66 years and older with incident COPD using multiple linked administrative health databases from the province of Ontario, Canada. Adherence to COPD medications was measured using time varying proportion of days covered based on insurance claims for medications dispensed at community pharmacies. The parametric g-formula was used to assess the association between time-varying adherence (in the last 90-days) to COPD medications and risk of COPD related hospitalizations and ED visits while accounting for time varying confounding by COPD severity.

RESULTS

Overall, 60,251 individuals with incident COPD were included; mean age was 76 (SD 7) and 59% were male. Mean adherence over the entire follow-up was 23% (SD 0.3). There were 7,248 (12%) COPD related ED visits (2.8 events per 100 person years [PY]) and 9,188 (15%) COPD related hospitalizations (3.5 events per 100 PY). Compared to those with 0% 90-day adherence, those with adherence between 1-33% had a 19% decreased risk of COPD related ED visits (adjusted risk ratio[aRR]:0.81, 95% confidence interval [CI]:0.78-0.83), those with adherence between 34%-67% had a 18% decreased risk (aRR: 0.82, 95% CI: 0.77-0.85) while those with 68%-100% 90-day adherence had a 63% increased risk of COPD related ED visits (aRR: 1.63, 95% CI: 1.47-1.78). Nearly identical results were obtained for COPD specific hospitalizations.

CONCLUSION

After accounting for time varying confounding by COPD severity, the highest time varying 90-days adherence was associated with an increased risk of both COPD related ED visits and hospitalizations compared to the lowest adherence categories. Differences in COPD severity between adherence categories, perception of need for medication management in the higher adherence categories, and potential residual confounding makes it difficult to disentangle the independent effects of adherence from the severity of the condition itself.

Major adverse cardiovascular events of enzalutamide versus abiraterone in prostate cancer: a retrospective cohort study

BACKGROUND

While the cardiovascular risks of androgen receptor pathway inhibitors have been studied, they were seldom compared directly. This study compares the risks of major adverse cardiovascular events (MACE) between enzalutamide and abiraterone among prostate cancer (PCa) patients.

METHODS

Adult PCa patients receiving either enzalutamide or abiraterone in addition to androgen deprivation therapy in Hong Kong between 1 December 1999 and 31 March 2021 were identified in this retrospective cohort study. Patients who switched between enzalutamide and abiraterone, initiated abiraterone used without steroids, or experienced prior cardiac events were excluded. Patients were followed-up until 30 September 2021. The primary outcomes were MACE, a composite of stroke, myocardial infarction (MI), Heart failure (HF), or all-cause mortality and a composite of adverse cardiovascular events (CACE) not including all-cause mortality. The secondary outcomes were individual components of MACE. Inverse probability treatment weighting was used to balance covariates between treatment groups.

RESULTS

In total, 1015 patients were analyzed (456 enzalutamide users and 559 abiraterone users; mean age 70.6 ± 8.8 years old) over a median follow-up duration of 11.3 (IQR: 5.3-21.3) months. Enzalutamide users had significantly lower risks of 4P-MACE (weighted hazard ratio (wHR) 0.71 [95% confidence interval (CI) 0.59-0.86], p < 0.001) and CACE (wHR 0.63 [95% CI: 0.42-0.96], p = 0.031), which remained consistent in multivariable analysis. Such an association may be stronger in patients aged ≥65 years or without diabetes mellitus and was independent of bilateral orchidectomy. Enzalutamide users also had significantly lower risks of MI (wHR 0.57 [95% CI: 0.33-0.97], p = 0.040) and all-cause mortality (wHR 0.71 [95% CI: 0.59-0.85], p < 0.001).

CONCLUSION

Enzalutamide was associated with lower cardiovascular risks than abiraterone in PCa patients.

Importance of accounting for timing of time-varying exposures in association studies: Hydrochlorothiazide and non-melanoma skin cancer

PURPOSE

Hydrochlorothiazide (HCTZ), a widely prescribed antihypertensive drug with photosensitising properties, has been linked with non-melanoma skin cancer (NMSC) risk. However, previous analyses did not fully explore if and how the impact of past HCTZ exposures accumulates with prolonged use and/or depends on time elapsed since exposures. Therefore, we used different models to more comprehensively assess how NMSC risk vary with HCTZ exposure, and explore how the results may depend on modeling strategies.

METHODS

We used different parametric models with alternative time-varying exposure metrics, and the flexible weighted cumulative exposure model (WCE) to estimate associations between HCTZ exposures and NMSC risk in a population-based cohort of HCTZ users over 65 years old, in the province of Ontario, Canada.

RESULTS

Among 3844 HCTZ users, 273 developed NMSC during up to 8 years of follow-up. In parametric models, based on all exposures, increased duration of past HCTZ use was associated with an increase of NMSC risk but cumulative dose showed no systematic association. Yet, WCE results suggested that only exposures taken 2.5-4 years in the past were associated with the current NMSC hazard. This finding led us to re-define the parametric models, which also confirmed that any HCTZ dose taken outside this time-window were not systematically associated with NMSC incidence.

CONCLUSIONS

Our analyses illustrate how flexible modeling may yield new insights into complex temporal relationships between a time-varying drug exposure and risks of adverse events. Duration and recency of antihypertensive agents exposures must be taken into account in evaluating risk and benefits.

Reporting, handling, and interpretation of time-varying drug treatments in observational studies using routinely collected healthcare data

BACKGROUND

Time-varying drug treatments are common in studies using routinely collected health data (RCD) for assessing treatment effects. This study aimed to examine how these studies reported, handled, and interpreted time-varying drug treatments.

METHODS

A systematic search was conducted on PubMed from 2018 to 2020. Eligible studies were those used RCD to explore drug treatment effects. We summarized the reporting characteristics and methods employed for handling time-varying treatments. Logistic regressions were performed to investigate the association between study characteristics and the reporting of time-varying treatments.

RESULTS

Two hundred and fifty-six studies were included, and 225 (87.9%) studies involved time-varying treatments. Of these, 24 (10.7%) reported the proportion of time-varying treatments and 105 (46.7%) reported methods used to handle time-varying treatments. Multivariable logistic regression showed that medical studies, prespecified protocol, and involvement of methodologists were associated with a higher likelihood of reporting the methods applied to handle time-varying treatments. Among the 105 studies that reported methods, as-treated analyses were the most commonly used analysis sets, which were employed in 73.9%, 75.3% and 88.2% of studies that reported approaches for treatment discontinuation, treatment switching and treatment add-on. Among the 225 studies involved time-varying treatments, 27 (12.0%) acknowledged the potential bias introduced by treatment change, of which 14 (51.9%) suggested that potential biases may impact acceptance or rejection of the null hypothesis.

CONCLUSIONS

Among observational studies using RCD, the underreporting about the presence and methods for handling time-varying treatments was largely common. The potential biases due to time-varying treatments have frequently been disregarded. Collaborative endeavors are strongly needed to enhance the prevailing practices.

Retrospective cohort study investigating synergism of air pollution and corticosteroid exposure in promoting cardiovascular and thromboembolic events in older adults

OBJECTIVE

To evaluate the synergistic effects created by fine particulate matter (PM2.5) and corticosteroid use on hospitalisation and mortality in older adults at high risk for cardiovascular thromboembolic events (CTEs).

DESIGN AND SETTING

A retrospective cohort study using a US nationwide administrative healthcare claims database.

PARTICIPANTS

A 50% random sample of participants with high-risk conditions for CTE from the 2008-2016 Medicare Fee-for-Service population.

EXPOSURES

Corticosteroid therapy and seasonal-average PM2.5.

MAIN OUTCOME MEASURES

Incidences of myocardial infarction or acute coronary syndrome (MI/ACS), ischaemic stroke or transient ischaemic attack, heart failure (HF), venous thromboembolism, atrial fibrillation and all-cause mortality. We assessed additive interactions between PM2.5 and corticosteroids using estimates of the relative excess risk due to interaction (RERI) obtained using marginal structural models for causal inference.

RESULTS

Among the 1 936 786 individuals in the high CTE risk cohort (mean age 76.8, 40.0% male, 87.4% white), the mean PM2.5 exposure level was 8.3±2.4 µg/m3 and 37.7% had at least one prescription for a systemic corticosteroid during follow-up. For all outcomes, we observed increases in risk associated with corticosteroid use and with increasing PM2.5 exposure. PM2.5 demonstrated a non-linear relationship with some outcomes. We also observed evidence of an interaction existing between corticosteroid use and PM2.5 for some CTEs. For an increase in PM2.5 from 8 μg/m3 to 12 μg/m3 (a policy-relevant change), the RERI of corticosteroid use and PM2.5 was significant for HF (15.6%, 95% CI 4.0%, 27.3%). Increasing PM2.5 from 5 μg/m3 to 10 μg/m3 yielded significant RERIs for incidences of HF (32.4; 95% CI 14.9%, 49.9%) and MI/ACSs (29.8%; 95% CI 5.5%, 54.0%).

CONCLUSION

PM2.5 and systemic corticosteroid use were independently associated with increases in CTE hospitalisations. We also found evidence of significant additive interactions between the two exposures for HF and MI/ACSs suggesting synergy between these two exposures.

Effects of Tanreqing Injection on ICU Mortality among ICU Patients Receiving Mechanical Ventilation: Time-Dependent Cox Regression Analysis of A Large Registry

OBJECTIVE

To assess whether the use of Tanreqing (TRQ) Injection could show improvements in time to extubation, intensive care unit (ICU) mortality, ventilator-associated events (VAEs) and infection-related ventilator associated complication (IVAC) among patients receiving mechanical ventilation (MV).

METHODS

A time-dependent cox-regression analysis was conducted using data from a well-established registry of healthcare-associated infections at ICUs in China. Patients receiving continuous MV for 3 days or more were included. A time-varying exposure definition was used for TRQ Injection, which were recorded on daily basis. The outcomes included time to extubation, ICU mortality, VAEs and IVAC. Time-dependent Cox models were used to compare the clinical outcomes between TRQ Injection and non-use, after controlling for the influence of comorbidities/conditions and other medications with both fixed and time-varying covariates. For the analyses of time to extubation and ICU mortality, Fine-Gray competing risk models were also used to measure competing risks and outcomes of interest.

RESULTS

Overall, 7,685 patients were included for the analyses of MV duration, and 7,273 patients for the analysis of ICU mortality. Compared to non-use, patients with TRQ Injection had a lower risk of ICU mortality (Hazards ratios (HR) 0.761, 95% CI, 0.581-0.997), and was associated with a higher hazard for time to extubation (HR 1.105, 95% CI, 1.005-1.216), suggesting a beneficial effect on shortened time to extubation. No significant differences were observed between TRQ Injection and non-use regarding VAEs (HR 1.057, 95% CI, 0.912-1.225) and IVAC (HR 1.177, 95% CI, 0.929-1.491). The effect estimates were robust when using alternative statistic models, applying alternative inclusion and exclusion criteria, and handling missing data by alternative approaches.

CONCLUSION

Our findings suggested that the use of TRQ Injection might lower mortality and improve time to extubation among patients receiving MV, even after controlling for the factor that the use of TRQ changed over time.

Metformin Use and Long-term Clinical Outcomes in Kidney Transplant Recipients

RATIONALE & OBJECTIVE

Metformin has been recommended for some patients with advanced chronic kidney disease. However, the value of metformin in kidney transplant recipients (KTRs) with pretransplant diabetes mellitus (DM) or posttransplant DM is uncertain. We investigated the clinical effects of metformin in KTRs.

STUDY DESIGN

Retrospective cohort study.

SETTING & PARTICIPANTS

A total of 1,995 KTRs with diabetes from 6 tertiary referral centers in the Republic of Korea.

EXPOSURE

Metformin usage was defined as the use of metformin for>90 days after kidney transplantation; 1,193 KTRs were metformin users, and 802 KTRs did not use metformin. Changing usage of metformin among those exposed for >90 days was also characterized.

OUTCOME

Primary outcomes were all-cause mortality and death-censored graft failure (DCGF). Secondary outcomes were biopsy-proven acute rejection (BPAR) and lactic acidosis events.

ANALYTICAL APPROACH

Survival analyses were conducted using multivariable Cox regression and competing risk analyses using Fine and Gray models. Changes in metformin use over time were modeled using a time-varying covariate. Metformin usage, mean daily dose, and hemoglobin A1c (HbA1c) changes were considered in the landmark analysis to address time-varying confounding.

RESULTS

Metformin use was associated with a lower risk of DCGF (adjusted hazard ratio [AHR], 0.47 [95% CI, 0.23-0.96], P=0.038); there was no significant association with all-cause mortality (AHR, 0.94 [95% CI, 0.32-2.76], P=0.915) or BPAR (AHR 0.98 [95% CI, 0.62-1.54], P=0.942). In the subgroup analysis, metformin usage was associated with a reduced risk of all-cause mortality and a lower risk of DCGF for both pretransplantation DM and posttransplant DM groups. Metformin usage was associated with a lower risk of BPAR in the posttransplant DM group, although it was less effective in the pretransplantation DM group. There was no confirmed case of metformin-associated lactic acidosis (MALA) in the present cohort. A higher dose of metformin was correlated with lower risks of DCGF and BPAR.

LIMITATIONS

Data on newer antidiabetic drugs such as SGLT2 inhibitors are limited, and there is potential limited generalizability to other populations.

CONCLUSIONS

Metformin usage may benefit KTRs, as evidenced by its association with a reduced risk of DCGF and the absence of MALA events. Randomized controlled trials are needed to validate these observational findings.

Reverse causation biases weighted cumulative exposure model estimates, but can be investigated in sensitivity analyses

OBJECTIVES

To examine the effects of reverse causation on estimates from the weighted cumulative exposure (WCE) model that is used in pharmacoepidemiology to explore drug-health outcome associations, and to identify sensitivity analyses for revealing such effects.

STUDY DESIGN AND SETTING

314,099 patients with diabetes under Clalit Health Services, Israel, were followed over 2002-2012. The association between metformin and pancreatic cancer (PC) was explored using a WCE model within the framework of discrete-time Cox regression. We used computer simulations to explore the effects of reverse causation on estimates of a WCE model and to examine sensitivity analyses for revealing and adjusting for reverse causation. We then applied those sensitivity analyses to our data.

RESULTS

Simulation demonstrated bias in the weighted cumulative exposure model and showed that sensitivity analysis could reveal and adjust for these biases. In our data, a positive association was observed (hazard ratio (HR) = 3.24, 95% confidence interval (CI): 2.24-4.73) with metformin exposure in the previous 2 years. After applying sensitivity analysis, assuming reverse causation operated up to 4 years before cancer diagnosis, the association between metformin and PC was no longer apparent.

CONCLUSION

Reverse causation can cause substantial bias in the WCE model. When suspected, sensitivity analyses based on causal analysis are advocated.

Comparison of Different Modeling Approaches for Prescription Opioid Use and Its Association With Adverse Events

Previous research linking opioid prescribing to adverse drug events has failed to properly account for the time-varying nature of opioid exposure. This study aimed to explore how the risk of opioid-related emergency department visits, readmissions, or deaths (composite outcome) varies with opioid dose and duration, comparing different novel modeling techniques. A prospective cohort of 1,511 hospitalized patients discharged from 2 McGill-affiliated hospitals in Montreal, 2014-2016, was followed from the first postdischarge opioid dispensation until 1 year after discharge. Marginal structural Cox proportional hazards models and their flexible extensions were used to explore the association between time-varying opioid use and the composite outcome. Weighted cumulative exposure models assessed cumulative effects of past use and explored how its impact depends on the recency of exposure. The patient mean age was 69.6 (standard deviation = 14.9) years; 57.7% were male. In marginal structural model analyses, current opioid use was associated with a 71% increase in the hazard of opioid-related adverse events (adjusted hazard ratio = 1.71, 95% confidence interval: 1.21, 2.43). The weighted cumulative exposure results suggested that the risk cumulates over the previous 50 days of opioid consumption. Flexible modeling techniques helped assess how the risk of opioid-related adverse events may be associated with time-varying opioid exposures while accounting for nonlinear relationships and the recency of past use.

Drug dosage modifications in 24 million in-patient prescriptions covering eight years: A Danish population-wide study of polypharmacy

Polypharmacy has generally been assessed by raw counts of different drugs administered concomitantly to the same patients; not with respect to the likelihood of dosage-adjustments. To address this aspect of polypharmacy, the objective of the present study was to identify co-medications associated with more frequent dosage adjustments. The data foundation was electronic health records from 3.2 million inpatient admissions at Danish hospitals (2008-2016). The likelihood of dosage-adjustments when two drugs were administered concomitantly were computed using Bayesian logistic regressions. We identified 3,993 co-medication pairs that associate significantly with dosage changes when administered together. Of these pairs, 2,412 (60%) did associate with readmission, mortality or longer stays, while 308 (8%) associated with reduced kidney function. In comparison to co-medications pairs that were previously classified as drug-drug interactions, pairs not classified as drug-drug interactions had higher odds ratios of dosage modifications than drug pairs with an established interaction. Drug pairs not corresponding to known drug-drug interactions while still being associated significantly with dosage changes were prescribed to fewer patients and mentioned more rarely together in the literature. We hypothesize that some of these pairs could be associated with yet to be discovered interactions as they may be harder to identify in smaller-scale studies.

Use of Tramadol vs Traditional Opioids and Adverse Outcomes in Patients with Inflammatory Bowel Disease: A Danish Nationwide Cohort Study

BACKGROUND

Use of traditional opioids (TOs) for pain management has been associated with adverse outcomes among patients with inflammatory bowel diseases (IBDs). It is unknown if similar associations exist for tramadol, a partial opioid agonist and serotonin and norephinephrine reuptake inhibitor. We sought to compare adverse outcomes associated with tramadol vs TOs in an IBD population.

METHODS

This nationwide cohort study included adults with IBD diagnosed from 1995 to 2021 in Denmark with subsequent prescriptions for tramadol or TOs. For each analgesic, 2 populations were assessed: initial users (first prescription) and persistent users (first 3 consecutive prescriptions within 365 days). Outcomes included infection, bowel obstruction/ileus, IBD surgery, and mortality within 90 days after the initial use index date (date of first prescription) and within 365 days after the persistent use index date (date of third prescription). Odds ratios adjusted for demographics, comorbidities, and IBD severity were calculated using multivariable logistic regression.

RESULTS

We identified 37 377 initial users and 15 237 persistent users of tramadol or TOs. Initial users of tramadol had lower adjusted odds of infection (adjusted odds ratio [OR], 0.80; 95% confidence interval [CI], 0.65-0.99), bowel obstruction/ileus (aOR, 0.74; 95% CI, 0.53-1.03), and mortality (aOR, 0.43; 95% CI, 0.35-0.55), and a higher adjusted odds of IBD-related surgery (aOR, 1.27; 95% CI, 1.02-1.60) vs initial users of TOs. Similar results were found for persistent users.

CONCLUSIONS

Tramadol was associated with lower odds of infection, bowel obstruction/ileus, and mortality vs TOs among patients with IBD. These associations may be impacted by residual confounding.

Asthma and COPD: Comparison with International Guidelines and Medication Adherence in Belgium

Asthma and chronic obstructive pulmonary disease (COPD) are major chronic conditions. It is possible to limit their impact by controlling symptoms, which limits exacerbations and worsening of the disease, by choosing the appropriate treatment and ensuring that the patient adheres to it. The main purpose of this study was to assess medication adherence and persistence with inhaled medications for chronic treatment of asthma and COPD, as well as to evaluate the factors influencing this adherence. Medication adherence was measured from January 2013 to December 2016 using continuous multiple-interval measures of medication availability (CMA). Persistence was evaluated by treatment episodes (TE). We analyzed the influence of different factors on CMA such as sex, age, type of device, and the realization of the "new medicines service" (NMS), introduced in Belgium in October 2013 to support patients in adhering to their treatment. We also analyzed the consumption of these inhaled medications within the Belgian population and compared them with the Global Initiative for Asthma (GINA) and the Global Initiative for Chronic Obstructive Lung Disease (GOLD) recommendations. Medication adherence varied greatly between the different pharmacological classes: inhaled corticosteroids (ICS) alone or in combination with long-acting beta agonists (LABA) had the lowest medication adherence and persistence, while adherence was highest for the long-acting muscarinic antagonists (LAMA) and LABA/LAMA associations. The NMS seemed to have a positive impact on medication adherence, although few patients completed the two guidance interviews offered by the service. In addition, only a minority of the targeted patients took advantage of this new service.

Survival benefits associated with being adherent and having longer persistence to adjuvant hormone therapy across up to five years among U.S. Medicare population with breast cancer

PURPOSE

To assess associations between adherence to and persistence with adjuvant hormone therapy and mortality among older women with breast cancer.

METHODS

The surveillance, epidemiology, and end results data linked with U.S. Medicare claims was used. This study included older women diagnosed with stage I-III hormone receptor-positive breast cancer from 2009 through 2017. Adherence was defined as having proportion of days covered (PDC) ≥ 0.80. Persistence was defined as having no discontinuation, i.e., no break of ≥ 180 continuous days. Length of persistence was calculated as time from therapy initiation to discontinuation. Cox models with time-dependent covariates were used to assess associations between adherence and persistence with mortality.

RESULTS

This study included 25,796 women. Adherence rates were 78.1 percent, 75.2 percent, 72.4 percent, 70.0 percent, and 61.5 percent from year 1 to year 5 after hormone therapy initiation. Persistence rates were 87.5 percent, 81.7 percent, 77.1 percent, 72.9 percent, and 68.9 percent through cumulative intervals of 1 year up to 5 years. Adherence was associated with all-cause mortality but not associated with breast cancer-specific mortality. Persistent women had lower risk of all-cause mortality and breast cancer-specific mortality. Each additional year of persistence had additional contributions to survival benefits (11% decreased risk of all-cause mortality and 37% decreased risk of breast cancer-specific mortality).

CONCLUSION

This study confirms the detrimental effect of nonadherence to adjuvant hormone therapy across up to 5 years on all-cause survival in older U.S. women. It also reveals the survival benefits associated with having longer persistence across up to 5 years.

"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.

Time-Varying Use of Proton Pump Inhibitors and Cognitive Impairment and Dementia: A Real-World Analysis from Germany

BACKGROUND

Cumulative evidence of dementia risk in patients taking proton pump inhibitors (PPIs) is still inconclusive, probably due to a variety of study designs.

OBJECTIVE

This study aimed to compare how the association between dementia risk and use of PPIs differs by different outcome and exposure definitions.

METHODS

We conceptualized a target trial using claims data with 7,696,127 individuals aged 40 years or older without previous dementia or mild cognitive impairment (MCI) from the Association of Statutory Health Insurance Physicians in Bavaria. Dementia was defined as either including or excluding MCI to compare how the results alter by different outcome definitions. We used weighted Cox models to estimate the PPI initiation effect on dementia risk and weighted pooled logistic regression to assess the effect of time-varying use versus non-use during 9 years of study period, including 1 year of wash-out period (2009-2018). The median follow-up time of PPI initiators and non-initiators was 5.4 and 5.8 years, respectively. We also evaluated the association between each PPI agent (omeprazole, pantoprazole, lansoprazole, esomeprazole, and combined use) and dementia risk.

RESULTS

A total of 105,220 (3.6%) PPI initiators and 74,697 (2.6%) non-initiators were diagnosed with dementia. Comparing PPI initiation with no initiation, the hazard ratio (HR) for dementia was 1.04 [95% confidence interval (CI) 1.03-1.05]. The HR for time-varying PPI use versus non-use was 1.85 (1.80-1.90). When MCI was included in the outcome, the number of outcomes increased to 121,922 in PPI initiators and 86,954 in non-initiators, but HRs remained similar, showing 1.04 (1.03-1.05) and 1.82 (1.77-1.86), respectively. Pantoprazole was the most frequently used PPI agent. Although the estimated HRs for the time-varying use effect of each PPI showed different ranges, all agents were associated with an increased dementia risk. A total of 105,220 (3.6%) PPI initiators and 74,697 (2.6%) non-initiators were diagnosed with dementia. Comparing PPI initiation with no initiation, the hazard ratio (HR) for dementia was 1.04 [95% confidence interval (CI) 1.03-1.05]. The HR for time-varying PPI use versus non-use was 1.85 (1.80-1.90). When MCI was included in the outcome, the number of outcomes increased to 121,922 in PPI initiators and 86,954 in non-initiators, but HRs remained similar, showing 1.04 (1.03-1.05) and 1.82 (1.77-1.86), respectively. Pantoprazole was the most frequently used PPI agent. Although the estimated HRs for the time-varying use effect of each PPI showed different ranges, all agents were associated with an increased dementia risk.A total of 105,220 (3.6%) PPI initiators and 74,697 (2.6%) non-initiators were diagnosed with dementia. Comparing PPI initiation with no initiation, the hazard ratio (HR) for dementia was 1.04 [95% confidence interval (CI) 1.03-1.05]. The HR for time-varying PPI use versus non-use was 1.85 (1.80-1.90). When MCI was included in the outcome, the number of outcomes increased to 121,922 in PPI initiators and 86,954 in non-initiators, but HRs remained similar, showing 1.04 (1.03-1.05) and 1.82 (1.77-1.86), respectively. Pantoprazole was the most frequently used PPI agent. Although the estimated HRs for the time-varying use effect of each PPI showed different ranges, all agents were associated with an increased dementia risk.A total of 105,220 (3.6%) PPI initiators and 74,697 (2.6%) non-initiators were diagnosed with dementia. Comparing PPI initiation with no initiation, the hazard ratio (HR) for dementia was 1.04 [95% confidence interval (CI) 1.03-1.05]. The HR for time-varying PPI use versus non-use was 1.85 (1.80-1.90). When MCI was included in the outcome, the number of outcomes increased to 121,922 in PPI initiators and 86,954 in non-initiators, but HRs remained similar, showing 1.04 (1.03-1.05) and 1.82 (1.77-1.86), respectively. Pantoprazole was the most frequently used PPI agent. Although the estimated HRs for the time-varying use effect of each PPI showed different ranges, all agents were associated with an increased dementia risk.A total of 105,220 (3.6%) PPI initiators and 74,697 (2.6%) non-initiators were diagnosed with dementia. Comparing PPI initiation with no initiation, the hazard ratio (HR) for dementia was 1.04 [95% confidence interval (CI) 1.03-1.05]. The HR for time-varying PPI use versus non-use was 1.85 (1.80-1.90). When MCI was included in the outcome, the number of outcomes increased to 121,922 in PPI initiators and 86,954 in non-initiators, but HRs remained similar, showing 1.04 (1.03-1.05) and 1.82 (1.77-1.86), respectively. Pantoprazole was the most frequently used PPI agent. Although the estimated HRs for the time-varying use effect of each PPI showed different ranges, all agents were associated with an increased dementia risk.A total of 105,220 (3.6%) PPI initiators and 74,697 (2.6%) non-initiators were diagnosed with dementia. Comparing PPI initiation with no initiation, the hazard ratio (HR) for dementia was 1.04 [95% confidence interval (CI) 1.03-1.05]. The HR for time-varying PPI use versus non-use was 1.85 (1.80-1.90). When MCI was included in the outcome, the number of outcomes increased to 121,922 in PPI initiators and 86,954 in non-initiators, but HRs remained similar, showing 1.04 (1.03-1.05) and 1.82 (1.77-1.86), respectively. Pantoprazole was the most frequently used PPI agent. Although the estimated HRs for the time-varying use effect of each PPI showed different ranges, all agents were associated with an increased dementia risk.A total of 105,220 (3.6%) PPI initiators and 74,697 (2.6%) non-initiators were diagnosed with dementia. Comparing PPI initiation with no initiation, the hazard ratio (HR) for dementia was 1.04 [95% confidence interval (CI) 1.03-1.05]. The HR for time-varying PPI use versus non-use was 1.85 (1.80-1.90). When MCI was included in the outcome, the number of outcomes increased to 121,922 in PPI initiators and 86,954 in non-initiators, but HRs remained similar, showing 1.04 (1.03-1.05) and 1.82 (1.77-1.86), respectively. Pantoprazole was the most frequently used PPI agent. Although the estimated HRs for the time-varying use effect of each PPI showed different ranges, all agents were associated with an increased dementia risk.A total of 105,220 (3.6%) PPI initiators and 74,697 (2.6%) non-initiators were diagnosed with dementia. Comparing PPI initiation with no initiation, the hazard ratio (HR) for dementia was 1.04 [95% confidence interval (CI) 1.03-1.05]. The HR for time-varying PPI use versus non-use was 1.85 (1.80-1.90). When MCI was included in the outcome, the number of outcomes increased to 121,922 in PPI initiators and 86,954 in non-initiators, but HRs remained similar, showing 1.04 (1.03-1.05) and 1.82 (1.77-1.86), respectively. Pantoprazole was the most frequently used PPI agent. Although the estimated HRs for the time-varying use effect of each PPI showed different ranges, all agents were associated with an increased dementia risk.A total of 105,220 (3.6%) PPI initiators and 74,697 (2.6%) non-initiators were diagnosed with dementia. Comparing PPI initiation with no initiation, the hazard ratio (HR) for dementia was 1.04 [95% confidence interval (CI) 1.03-1.05]. The HR for time-varying PPI use versus non-use was 1.85 (1.80-1.90). When MCI was included in the outcome, the number of outcomes increased to 121,922 in PPI initiators and 86,954 in non-initiators, but HRs remained similar, showing 1.04 (1.03-1.05) and 1.82 (1.77-1.86), respectively. Pantoprazole was the most frequently used PPI agent. Although the estimated HRs for the time-varying use effect of each PPI showed different ranges, all agents were associated with an increased dementia risk.

CONCLUSION

Our large study supports existing evidence that PPI use is related to an increased risk of dementia.

Target Trial Emulation Using Hospital-Based Observational Data: Demonstration and Application in COVID-19

Methodological biases are common in observational studies evaluating treatment effectiveness. The objective of this study is to emulate a target trial in a competing risks setting using hospital-based observational data. We extend established methodology accounting for immortal time bias and time-fixed confounding biases to a setting where no survival information beyond hospital discharge is available: a condition common to coronavirus disease 2019 (COVID-19) research data. This exemplary study includes a cohort of 618 hospitalized patients with COVID-19. We describe methodological opportunities and challenges that cannot be overcome applying traditional statistical methods. We demonstrate the practical implementation of this trial emulation approach via clone-censor-weight techniques. We undertake a competing risk analysis, reporting the cause-specific cumulative hazards and cumulative incidence probabilities. Our analysis demonstrates that a target trial emulation framework can be extended to account for competing risks in COVID-19 hospital studies. In our analysis, we avoid immortal time bias, time-fixed confounding bias, and competing risks bias simultaneously. Choosing the length of the grace period is justified from a clinical perspective and has an important advantage in ensuring reliable results. This extended trial emulation with the competing risk analysis enables an unbiased estimation of treatment effects, along with the ability to interpret the effectiveness of treatment on all clinically important outcomes.

A Method to Create Directed Acyclic Graphs from Cycles of Transmission of Zoonotic and Vector-Borne Infectious Agents

Background: The life cycles of zoonotic and vector-borne diseases can be complex. This complexity makes it challenging to identify factors that confound the association between an exposure of interest and infection in one of the susceptible hosts. In epidemiology, directed acyclic graphs (DAGs) can be used to visualize the relationships between exposures and outcomes and also to identify which factors confound the association between exposure and the outcome of interest. However, DAGs can only be used in situations where no cycle exists in the causal relationships being represented. This is problematic for infectious agents that cycle between hosts. Zoonoses and vector-borne diseases pose additional challenges with DAG construction since multiple required or optional hosts of different species may be part of the cycle. Methods: We review the existing examples of DAGs created for nonzoonotic infectious agents. We then demonstrate how to cut the transmission cycle to create DAGs where infection of a specific host species is the outcome of interest. We adapt our method to create DAGs using examples of transmission and host characteristics common to many zoonotic and vector-borne infectious agents. Results: We demonstrate our method using the transmission cycle of West Nile virus to create a simple transmission DAG that lacks a cycle. Conclusions: Using our work, investigators can create DAGs to help identify confounders of the relationships between modifiable risk factors and infection. Ultimately, a better understanding and control of confounding in measuring the impact of such risk factors can be used to inform health policy, guide public health and animal health interventions, and uncover gaps needing further research attention.

Comparison of cohort and nested case-control designs for estimating the effect of time-varying drug exposure on the risk of adverse event in the presence of ties

Cohort and nested case-control (NCC) designs are frequently used in pharmacoepidemiology to assess the associations of drug exposure that can vary over time with the risk of an adverse event. Although it is typically expected that estimates from NCC analyses are similar to those from the full cohort analysis, with moderate loss of precision, only few studies have actually compared their respective performance for estimating the effects of time-varying exposures (TVE). We used simulations to compare the properties of the resulting estimators of these designs for both time-invariant exposure and TVE. We varied exposure prevalence, proportion of subjects experiencing the event, hazard ratio, and control-to-case ratio and considered matching on confounders. Using both designs, we also estimated the real-world associations of time-invariant ever use of menopausal hormone therapy (MHT) at baseline and updated, time-varying MHT use with breast cancer incidence. In all simulated scenarios, the cohort-based estimates had small relative bias and greater precision than the NCC design. NCC estimates displayed bias to the null that decreased with a greater number of controls per case. This bias markedly increased with higher proportion of events. Bias was seen with Breslow's and Efron's approximations for handling tied event times but was greatly reduced with the exact method or when NCC analyses were matched on confounders. When analyzing the MHT-breast cancer association, differences between the two designs were consistent with simulated data. Once ties were taken correctly into account, NCC estimates were very similar to those of the full cohort analysis.

Clinical outcomes following discontinuation of renin-angiotensin-system inhibitors in patients with type 2 diabetes and advanced chronic kidney disease: A prospective cohort study

BACKGROUND

Renin-angiotensin-system inhibitors (RASi), that include angiotensin converting enzyme inhibitors (ACEis) and angiotensin receptor blockers (ARBs) reduce proteinuria, delay chronic kidney disease (CKD) progression, protect against cardiovascular events and heart failure hospitalizations. We examined the associations of discontinuation of ACEi/ARBs with risk of clinical outcomes in Chinese patients with type 2 diabetes (T2D) and advanced-CKD (estimated-glomerular filtration rate [eGFR] <30 ml/min/1.73 m²).

METHODS

We conducted a prospective, population-based cohort study including 10,400 patients with T2D in Hong Kong stratified by continuation of ACEi/ARBs within 6 months after reaching eGFR <30 ml/min/1.73 m² from January 01, 2002 to December 31, 2018 and observed until December 31, 2019. The primary outcomes were death, major-adverse cardiovascular events (MACE), heart failure, end-stage kidney disease (ESKD), and all-cause mortality. Cox-model with time-dependent exposure and covariates was used to estimate the hazard ratio (HR) of outcomes in a propensity-score overlap-weighted cohort. The risk of occurrence of hyperkalemia (plasma potassium >5.5 mmol/L) in discontinued-ACEi/ARBs versus continued-ACEi/ARBs users was assessed in a register-based cohort.

FINDINGS

In the population-based cohort of 10,400 ACEi/ARBs users with new-onset eGFR<30 ml/min/1.73 m², 1766 (17.0%) discontinued ACEi/ARBs and 8634 (83.0%) persisted with treatment. During a median follow-up of 3.6 (interquartile range, IQR: 2.11-5.8) years (41,623 person-years), 13.5%, 12.9%, and 27.6% had incident MACE, heart failure and ESKD respectively, and 35.8% died. Discontinued-ACEi/ARBs use was associated with higher risk of MACE (HR = 1.27, 95% CI: 1.08-1.49), heart failure (HR = 1.85, 95% CI: 1.53-2.25) and ESKD (HR = 1.30, 95% CI: 1.17-1.43), and neutral risk of all-cause mortality (HR = 0.93, 95% CI: 0.86-1.01) compared to counterparts with continued use. In the register-based cohort (583 discontinued-ACEi/ARBs users and 3817 continued-ACEi/ARBs users), discontinued-ACEi/ARBs had neutral risk of hyperkalemia (HR = 0.95, 95% CI: 0.84-1.08).

INTERPRETATION

Discontinuation of ACEi/ARBs was associated with increased risk of cardiovascular-renal events supporting their continued use in patients with T2D and advanced-CKD.

FUNDING

CUHK Impact Research Fellowship Scheme.

Key considerations for choosing a statistical method to deal with incomplete treatment adherence in pragmatic trials

Pragmatic trials offer practical means of obtaining real-world evidence to help improve decision-making in comparative effectiveness settings. Unfortunately, incomplete adherence is a common problem in pragmatic trials. The commonly used methods in randomized control trials often cannot handle the added complexity imposed by incomplete adherence, resulting in biased estimates. Several naive methods and advanced causal inference methods (e.g., inverse probability weighting and instrumental variable-based approaches) have been used in the literature to deal with incomplete adherence. Practitioners and applied researchers are often confused about which method to consider under a given setting. This current work is aimed to review commonly used statistical methods to deal with non-adherence along with their key assumptions, advantages, and limitations, with a particular focus on pragmatic trials. We have listed the applicable settings for these methods and provided a summary of available software. All methods were applied to two hypothetical datasets to demonstrate how these methods perform in a given scenario, along with the R codes. The key considerations include the type of intervention strategy (point treatment settings, where treatment is administered only once versus sustained treatment settings, where treatment has to be continued over time) and availability of data (e.g., the extent of measured or unmeasured covariates that are associated with adherence, dependent confounding impacted by past treatment, and potential violation of assumptions). This study will guide practitioners and applied researchers to use the appropriate statistical method to address incomplete adherence in pragmatic trial settings for both the point and sustained treatment strategies.

Association between malignancy risk and Janus kinase inhibitors versus tumour necrosis factor inhibitors in Korean patients with rheumatoid arthritis: a nationwide population-based study

OBJECTIVE

To determine the risk of malignancy in Korean patients with rheumatoid arthritis (RA) receiving Janus kinase inhibitors (JAKis) compared with tumour necrosis factor inhibitors (TNFis).

METHODS

A retrospective cohort of patients with RA initiating their first JAKi or TNFi was established using the Korean National Health Insurance database between 2015 and 2019. They were followed up from treatment initiation to the occurrence of malignancy, drug discontinuation, death or until December 2019. Baseline features of the patients were balanced through inverse probability of treatment weighting (IPTW) using a propensity score. A Cox proportional hazard model was established to estimate the HR for malignancy risk in JAKi users compared with TNFi users.

RESULTS

A total of 4929 patients (1064 JAKi-treated and 3865 TNFi-treated patients) were included, and the observation periods were 1288.6 person-years (PYs) for JAKi users and 6823.8 PYs for TNFi users. The incidence rates of overall malignancy were 0.54 per 100 PYs (95% CI 0.26 to 1.14) in JAKi users and 0.85 per 100 PYs (95% CI 0.66 to 1.10) in TNFi users. In IPTW analysis with a balanced sample (4101 JAKi-treated and 5131 TNFi-treated patients), HR was 0.83 (95% CI 0.55 to 1.27) for overall malignancy: 0.77 (95% CI 0.50 to 1.19) for solid malignancy and 2.86 (95% CI 0.41 to 20.00) for haematological malignancy.

CONCLUSION

Malignancy risk in Korean patients with RA was not increased with JAKi use compared with TNFi use.

Management of type 2 diabetes with a treat-to-benefit approach improved long-term cardiovascular outcomes under routine care

BACKGROUND

Results of cardiovascular outcome trials enabled a shift from "treat-to-target" to "treat-to-benefit" paradigm in the management of type 2 diabetes (T2D). However, studies validating such approach are limited. Here, we examined whether treatment according to international recommendations for the pharmacological management of T2D had an impact on long-term outcomes.

METHODS

This was an observational study conducted on outpatient data collected in 2008-2018 (i.e. prior to the "treat-to-benefit" shift). We defined 6 domains of treatment based on the ADA/EASD consensus covering all disease stages: first- and second-line treatment, intensification, use of insulin, cardioprotective, and weight-affecting drugs. At each visit, patients were included in Group 1 if at least one domain deviated from recommendation or in Group 2 if aligned with recommendations. We used Cox proportional hazard models with time-dependent co-variates or Cox marginal structural models (with inverse-probability of treatment weighing evaluated at each visit) to adjust for confounding factors and evaluate three outcomes: major adverse cardiovascular events (MACE), hospitalization for heart failure or cardiovascular mortality (HF-CVM), and all-cause mortality.

RESULTS

We included 5419 patients, on average 66-year old, 41% women, with a baseline diabetes duration of 7.6 years. Only 11.7% had pre-existing cardiovascular disease. During a median follow-up of 7.3 years, patients were seen 12 times at the clinic, and we recorded 1325 MACE, 1593 HF-CVM, and 917 deaths. By the end of the study, each patient spent on average 63.6% of time in Group 1. In the fully adjusted model, being always in Group 2 was associated with a 45% lower risk of MACE (HR 0.55; 95% C.I. 0.46-0.66; p < 0.0001) as compared to being in Group 1. The corresponding HF-CVM and mortality risk were similar (HR 0.56; 95%CI 0.47-0.66, p < 0.0001 and HR 0.56; 95% C.I. 0.45-0.70; p < 0.0001. respectively). Sensitivity analyses confirmed these results. No single domain individually explained the better outcome of Group 2, which remained significant in all subgroups.

CONCLUSION

Managing patients with T2D according to a "treat-to-benefit" approach based international standards was associated with a lower risk of MACE, heart failure, and mortality. These data provide ex-post validation of the ADA/EASD treatment algorithm.

Discontinuation and tapering of prescribed opioids and risk of overdose among people on long-term opioid therapy for pain with and without opioid use disorder in British Columbia, Canada: A retrospective cohort study

BACKGROUND

The overdose crisis in North America has prompted system-level efforts to restrict opioid prescribing for chronic pain. However, little is known about how discontinuing or tapering prescribed opioids for chronic pain shapes overdose risk, including possible differential effects among people with and without concurrent opioid use disorder (OUD). We examined associations between discontinuation and tapering of prescribed opioids and risk of overdose among people on long-term opioid therapy for pain, stratified by diagnosed OUD and prescribed opioid agonist therapy (OAT) status.

METHODS AND FINDINGS

For this retrospective cohort study, we used a 20% random sample of residents in the provincial health insurance client roster in British Columbia (BC), Canada, contained in the BC Provincial Overdose Cohort. The study sample included persons aged 14 to 74 years on long-term opioid therapy for pain (≥90 days with ≥90% of days on therapy) between October 2014 and June 2018 (n = 14,037). At baseline, 7,256 (51.7%) persons were female, the median age was 55 years (quartile 1-3: 47-63), 227 (1.6%) persons had been diagnosed with OUD (in the past 3 years) and recently (i.e., in the past 90 days) been prescribed OAT, and 483 (3.4%) had been diagnosed with OUD but not recently prescribed OAT. The median follow-up duration per person was 3.7 years (quartile 1-3: 2.6-4.0). Marginal structural Cox regression with inverse probability of treatment weighting (IPTW) was used to estimate the effect of prescribed opioid treatment for pain status (discontinuation versus tapered therapy versus continued therapy [reference]) on risk of overdose (fatal or nonfatal), stratified by the following groups: people without diagnosed OUD, people with diagnosed OUD receiving OAT, and people with diagnosed OUD not receiving OAT. In marginal structural models with IPTW adjusted for a range of demographic, prescription, comorbidity, and social-structural exposures, discontinuing opioids (i.e., ≥7-day gap[s] in therapy) was associated with increased overdose risk among people without OUD (adjusted hazard ratio [AHR] = 1.44; 95% confidence interval [CI] 1.12, 1.83; p = 0.004), people with OUD not receiving OAT (AHR = 3.18; 95% CI 1.87, 5.40; p < 0.001), and people with OUD receiving OAT (AHR = 2.52; 95% CI 1.68, 3.78; p < 0.001). Opioid tapering (i.e., ≥2 sequential decreases of ≥5% in average daily morphine milligram equivalents) was associated with decreased overdose risk among people with OUD not receiving OAT (AHR = 0.31; 95% CI 0.14, 0.67; p = 0.003). The main study limitations are that the outcome measure did not capture overdose events that did not result in a healthcare encounter or death, medication dispensation may not reflect medication adherence, residual confounding may have influenced findings, and findings may not be generalizable to persons on opioid therapy in other settings.

CONCLUSIONS

Discontinuing prescribed opioids was associated with increased overdose risk, particularly among people with OUD. Prescribed opioid tapering was associated with reduced overdose risk among people with OUD not receiving OAT. These findings highlight the need to avoid abrupt discontinuation of opioids for pain. Enhanced guidance is needed to support prescribers in implementing opioid therapy tapering strategies with consideration of OUD and OAT status.

Adapting SIMEX to correct for bias due to interval-censored outcomes in survival analysis with time-varying exposure

Many clinical and epidemiological applications of survival analysis focus on interval-censored events that can be ascertained only at discrete times of clinic visits. This implies that the values of time-varying covariates are not correctly aligned with the true, unknown event times, inducing a bias in the estimated associations. To address this issue, we adapted the simulation-extrapolation (SIMEX) methodology, based on assessing how the estimates change with the artificially increased time between clinic visits. We propose diagnostics to choose the extrapolating function. In simulations, the SIMEX-corrected estimates reduced considerably the bias to the null and generally yielded a better bias/variance trade-off than conventional estimates. In a real-life pharmacoepidemiological application, the proposed method increased by 27% the excess hazard of the estimated association between a time-varying exposure, representing the 2-year cumulative duration of past use of a hypertensive medication, and the hazard of nonmelanoma skin cancer (interval-censored events). These simulation-based and real-life results suggest that the proposed SIMEX-based correction may help improve the accuracy of estimated associations between time-varying exposures and the hazard of interval-censored events in large cohort studies where the events are recorded only at relatively sparse times of clinic visits/assessments. However, these advantages may be less certain for smaller studies and/or weak associations.

Antithrombotic treatment switching in elderly patients with atrial fibrillation and the risk of thromboembolism, bleeding, and cardiac death

Background

Risks of antithrombotic switching is not investigated in elderly atrial fibrillation patients.

Objectives

To investigate the effectiveness and safety of antithrombotic treatment and switching of antithrombotic treatment in elderly patients (aged 75 years or older) with atrial fibrillation (AF).

Methods

We conducted a cohort study of 2943 patients with AF (Carrebean‐elderly), hospitalized during 2010–2017. Cox models were used to estimate the association of antithrombotic treatment (warfarin, direct oral anticoagulants [DOAC] and non–guideline‐recommended therapy [NG], i.e., aspirin and low‐molecular‐weight heparin) at discharge and antithrombotic treatment switching during follow‐up with the risk of a composite and single end points of thromboembolism, bleeding, and cardiac death. Crude and adjusted risk estimates were expressed as hazard ratios (HRs) with 95% confidence intervals (CIs). All‐cause death was evaluated, with competing risk regression and estimates expressed as subhazard ratios and 95% CIs.

Results

We observed an increased risk for the composite end point associated with NG as compared to warfarin at discharge (HR, 1.18; 95% CI, 1.01–1.38) with congruent competing risk regression results, while no significant risk difference was seen for DOACs compared to warfarin (HR, 1.12; 95% CI, 0.92–1.36). Switching from NG to warfarin/DOAC and from warfarin to DOAC occurred in 30.4% and 33.1% of respective antithrombotic treatment groups at discharge and was associated with a decreased risk for the composite end point with an adjusted HR of 0.45 (95% CI, 0.32–0.63) and a HR of 0.50 (95% CI, 0.38–0.65), respectively.

Conclusions

Antithrombotic treatment switching is common in the elderly AF population. Importantly, switching to guideline‐recommended treatment has a favorable impact on both effectiveness and safety.

Effects of RAS inhibitors on all-site cancers and mortality in the Hong Kong diabetes surveillance database (2002-2019)

Background

Cancer is replacing cardiovascular-disease as a leading cause of death in type 2 diabetes (T2D). The association of RAS-inhibitors (RASi) and cancer, including differences between angiotensin-converting-enzyme-inhibitor (ACEi) and angiotensin-receptor-blocker (ARBs) as well as their associations independent of blood pressure lowering, remains inconclusive in T2D.

Methods

We conducted a cohort study with new-user design in 253,491 patients in the Hong-Kong-Diabetes-Surveillance-Database (HKDSD) in 2002-2019. We evaluated the associations of time-varying RASi use (ACEi and ARBs) with all-site cancer, diabetes-related cancers, and cancer-specific mortality including comparison with new-users of calcium-channel-blockers (CCBs) as an active-comparator group.

Findings

Of 253,491, 133,730 (52.8%) were new-RASi and 119,761 (47.2%) were non-RASi users with a median follow-up period of 6.3 (interquartile ragne: 3.4-9.2) years (1,678,719 patient-years). After propensity-score weighting and adjustment for time-varying covariables, RASi use was associated with lower risk of all-site cancer (HR=0.76, 95%CI: 0.74-0.79), diabetes-related cancer (HR=0.79, 95%CI: 0.75-0.84), cancer-specific mortality (HR=0.50, 95%CI: 0.47-0.53), and diabetes-related cancer mortality (HR=0.49, 95%CI: 0.45-0.54) versus non-RASi. Amongst RASi users, ARBs use was associated with lower risk of cancer-specific mortality versus ACEi (HR=0.77, 95%CI: 0.66-0.91). Use of RASi was associated with an estimated-prevention of 2.6 (95%CI: 2.3-3.0) all-site cancer per-1000-person-years and 2.2 (95%CI: 2.0-2.5) cancer-related mortality per-1000-person-years. Lower risk of cancer-specific mortality was similarly observed in new-RASi compared with new-CCBs users.

Interpretation

RASi use was independently associated with lower cancer risk in T2D with stronger associations in users of ARBs than ACEi. The benefits of RASi in patients with diabetes might go beyond cardiovascular-renal protection if confirmed by other real-world studies and trials.

Funding

Dr. Aimin Yang was supported by a CUHK Impact-Research-Fellowship Scheme.

Breastfeeding, pregnancy, medicines, neurodevelopment, and population databases: the information desert

Background

The pharmacoepidemiology of the long-term benefits and harms of medicines in pregnancy and breastfeeding has received little attention. The impact of maternal medicines on children is increasingly recognised as a source of avoidable harm. The focus of attention has expanded from congenital anomalies to include less visible, but equally important, outcomes, including cognition, neurodevelopmental disorders, educational performance, and childhood ill-health. Breastfeeding, whether as a source of medicine exposure, a mitigator of adverse effects or as an outcome, has been all but ignored in pharmacoepidemiology and pharmacovigilance: a significant ‘blind spot’.

Whole-population data on breastfeeding: why we need them

Optimal child development and maternal health necessitate breastfeeding, yet little information exists to guide families regarding the safety of medicine use during lactation. Breastfeeding initiation or success may be altered by medicine use, and breastfeeding may obscure the true relationship between medicine exposure during pregnancy and developmental outcomes. Absent or poorly standardised recording of breastfeeding in most population databases hampers analysis and understanding of the complex relationships between medicine, pregnancy, breastfeeding and infant and maternal health. The purpose of this paper is to present the arguments for breastfeeding to be included alongside medicine use and neurodevelopmental outcomes in whole-population database investigations of the harms and benefits of medicines during pregnancy, the puerperium and postnatal period. We review: 1) the current situation, 2) how these complexities might be accommodated in pharmacoepidemiological models, using antidepressants and antiepileptics as examples; 3) the challenges in obtaining comprehensive data.

Conclusions

The scarcity of whole-population data and the complexities of the inter-relationships between breastfeeding, medicines, co-exposures and infant outcomes are significant barriers to full characterisation of the benefits and harms of medicines during pregnancy and breastfeeding. This makes it difficult to answer the questions: ‘is it safe to breastfeed whilst taking this medicine’, and ‘will this medicine interfere with breastfeeding and/ or infants’ development’?

Impact of time-varying exposure on estimated effects in observational studies using routinely collected data: protocol for a cross-sectional study

INTRODUCTION

Time-varying exposure is an important issue that should be addressed in longitudinal observational studies using routinely collected data (RCD) for drug treatment effects. How well investigators designed, analysed and reported time-varying exposure, and to what extent the divergence that can be observed between different methods used for handling time-varying exposure in these studies remains uncertain. We will conduct a cross-sectional study to comprehensively address this question.

METHODS AND ANALYSIS

We have developed a comprehensive search strategy to identify all studies exploring drug treatment effects including both effectiveness and safety that used RCD and were published in core journals between 2018 and 2020. We will collect information regarding general study characteristics, data source profile, methods for handling time-varying exposure, results and the interpretation of findings from each eligibility. Paired reviewers will screen and extract data, resolving disagreements through discussion. We will describe the characteristics of included studies, and summarise the method used for handling time-varying exposure in primary analysis and sensitivity analysis. We will also compare the divergence between different approaches for handling time-varying exposure using ratio of risk ratios.

ETHICS AND DISSEMINATION

No ethical approval is required because the data we will use do not include individual patient data. Findings will be disseminated through peer-reviewed publications.

Statins was not associated with hepatocellular carcinoma after controlling for time-varying confounders in patients with diabetes

OBJECTIVE

We examined the association between statin use and hepatocellular carcinoma (HCC) incidence in patients with diabetes using marginal structural models (MSMs) estimated by inverse probability weight (IPW), which adjusts for time-varying confounders that are also mediators, and we compared the results with conventional regression methods.

STUDY DESIGN AND SETTING

This retrospective cohort study included 245,122 patients with type 2 diabetes who were new users of lipid-lowering drugs identified using the claims data of a universal health insurance program. Statin exposure was time-updated every 3 months during the follow-up period. Stabilized IPW was calculated and accounted for chronic liver diseases considering as time-dependent confounders affected by past statin exposure.

RESULTS

Over a median follow-up of 5.2 years, 1,694 patients developed HCC. In the conventional regression analysis, the hazard ratio of HCC associated with statin use was 0.88 (95% CI: 0.79-0.97) after adjusting for baseline covariates and 0.97 (95% CI: 0.87-1.08) after additionally adjusting for time-varying covariates. The hazard ratio increased to 1.11 (95% CI: 0.94-1.31) using the MSM approach.

CONCLUSION

Statins use was not associated with the risk of developing HCC in patients with diabetes. Our findings highlight the importance of controlling time-varying confounders in observational studies.

Alcohol brief intervention, specialty treatment and drinking outcomes at 12 months: Results from a systematic alcohol screening and brief intervention initiative in adult primary care

BACKGROUND

Alcohol screening, brief intervention and referral to treatment (SBIRT) in adult primary care is an evidence-based, public health strategy to address unhealthy alcohol use, but evidence of effectiveness of alcohol brief intervention (ABI) in real-world implementation is lacking.

METHODS

We fit marginal structural models with inverse probability weighting to estimate the causal effects of ABI on 12-month drinking outcomes using longitudinal electronic health records data for 312,056 adults with a positive screening result for unhealthy drinking between 2014 and 2017 in a large healthcare system that implemented systematic primary care-based SBIRT. We examined effects of ABI with and without adjusting for receipt of specialty alcohol use disorder (AUD) treatment, and whether effects varied by patient demographic characteristics and alcohol use patterns.

RESULTS

Receiving ABI resulted in significantly greater reductions in heavy drinking days (mean difference [95% CI] = -0.26 [-0.45, -0.08]), drinking days per week (-0.04 [-0.07, -0.01]), drinks per drinking day (-0.05 [-0.08, -0.02]) and drinks per week (-0.16 [-0.27, -0.04]). Effects of ABI on 12-month drinking outcomes varied by baseline consumption level, age group and whether patients already had an AUD, with better improvement in those who were drinking at levels exceeding only daily limits, younger, and without an AUD.

CONCLUSIONS

Systematic ABI in adult primary care has the potential to reduce drinking among people with unhealthy drinking considerably on both an individual and population level. More research is needed to help optimize ABI, in particular tailoring it to diverse sub-populations, and studying its long-term public health impact.

Rationale and performances of a data-driven method for computing the duration of pharmacological prescriptions using secondary data sources

The assessment of the duration of pharmacological prescriptions is an important phase in pharmacoepidemiologic studies aiming to investigate persistence, effectiveness or safety of treatments. The Sessa Empirical Estimator (SEE) is a new data-driven method which uses k-means algorithm for computing the duration of pharmacological prescriptions in secondary data sources when this information is missing or incomplete. The SEE was used to compute durations of exposure to pharmacological treatments where simulated and real-world data were used to assess its properties comparing the exposure status extrapolated with the method with the “true” exposure status available in the simulated and real-world data. Finally, the SEE was also compared to a Researcher-Defined Duration (RDD) method. When using simulated data, the SEE showed accuracy of 96% and sensitivity of 96%, while when using real-world data, the method showed sensitivity ranging from 78.0 (nortriptyline) to 95.1% (propafenone). When compared to the RDD, the method had a lower median sensitivity of 2.29% (interquartile range 1.21–4.11%). The SEE showed good properties and may represent a promising tool to assess exposure status when information on treatment duration is not available.

Confounding adjustment methods in longitudinal observational data with a time-varying treatment: a mapping review

OBJECTIVES

To adjust for confounding in observational data, researchers use propensity score matching (PSM), but more advanced methods might be required when dealing with longitudinal data and time-varying treatments as PSM might not include possible changes that occurred over time. This study aims to explore which confounding adjustment methods have been used in longitudinal observational data to estimate a treatment effect and identify potential inappropriate use of PSM.

DESIGN

Mapping review.

DATA SOURCES

We searched PubMed, from inception up to January 2021, for studies in which a treatment was evaluated using longitudinal observational data.

ELIGIBILITY CRITERIA

Methodological, non-medical and cost-effectiveness papers were excluded, as were non-English studies and studies that did not study a treatment effect.

DATA EXTRACTION AND SYNTHESIS

Studies were categorised based on time of treatment: at baseline (interventions performed at start of follow-up) or time-varying (interventions received asynchronously during follow-up) and sorted based on publication year, time of treatment and confounding adjustment method. Cumulative time series plots were used to investigate the use of different methods over time. No risk-of-bias assessment was performed as it was not applicable.

RESULTS

In total, 764 studies were included that met the eligibility criteria. PSM (165/201, 82%) and inverse probability weighting (IPW; 154/502, 31%) were most common for studies with a treatment at baseline (n=201) and time-varying treatment (n=502), respectively. Of the 502 studies with a time-varying treatment, 123 (25%) used PSM with baseline covariates, which might be inappropriate. In the past 5 years, the proportion of studies with a time-varying treatment that used PSM over IPW increased.

CONCLUSIONS

PSM is the most frequently used method to correct for confounding in longitudinal observational data. In studies with a time-varying treatment, PSM was potentially inappropriately used in 25% of studies. Confounding adjustment methods designed to deal with a time-varying treatment and time-varying confounding are available, but were only used in 45% of the studies with a time-varying treatment.

Assessing proximate intermediates between ambient temperature, hospital admissions, and mortality in hemodialysis patients

BACKGROUND

Typical thermoregulatory responses to elevated temperatures among healthy individuals include reduced blood pressure and perspiration. Individuals with end-stage kidney disease (ESKD) are susceptible to systemic fluctuations caused by ambient temperature changes that may increase morbidity and mortality. We investigated whether pre-dialysis systolic blood pressure (preSBP) and interdialytic weight gain (IDWG) can independently mediate the association between ambient temperature, all-cause hospital admissions (ACHA), and all-cause mortality (ACM).

METHODS

The study population consisted of ESKD patients receiving hemodialysis treatments at Fresenius Medical Care facilities in Philadelphia County, PA, from 2011 to 2019 (n = 1981). Within a time-to-event framework, we estimated the association between daily maximum dry-bulb temperature (TMAX) and, as separate models, ACHA and ACM during warmer calendar months. Clinically measured preSBP and IDWG responses to temperature increases were estimated using linear mixed effect models. We employed the difference (c-c') method to decompose total effect models for ACHA and ACM using preSBP and IDWG as time-dependent mediators. Covariate adjustments for exposure-mediator and total and direct effect models include age, race, ethnicity, blood pressure medication use, treatment location, preSBP, and IDWG. We considered lags up to two days for exposure and 1-day lag for mediator variables (Lag 2-Lag 1) to assure temporality between exposure-outcome models. Sensitivity analyses for 2-day (Lag 2-only) and 1-day (Lag 1-only) lag structures were also conducted.

RESULTS

Based on Lag 2- Lag 1 temporal ordering, 1 °C increase in daily TMAX was associated with increased hazard of ACHA by 1.4% (adjusted hazard ratio (HR), 1.014; 95% confidence interval, 1.007-1.021) and ACM 7.5% (adjusted HR, 1.075, 1.050-1.100). Short-term lag exposures to 1 °C increase in temperature predicted mean reductions in IDWG and preSBP by 0.013-0.015% and 0.168-0.229 mmHg, respectively. Mediation analysis for ACHA identified significant indirect effects for all three studied pathways (preSBP, IDWG, and preSBP + IDWG) and significant indirect effects for IDWG and conjoined preSBP + IDWG pathways for ACM. Of note, only 1.03% of the association between temperature and ACM was mediated through preSBP. The mechanistic path for IDWG, independent of preSBP, demonstrated inconsistent mediation and, consequently, potential suppression effects in ACHA (-15.5%) and ACM (-6.3%) based on combined pathway models. Proportion mediated estimates from preSBP + IDWG pathways achieved 2.2% and 0.3% in combined pathway analysis for ACHA and ACM outcomes, respectively. Lag 2 discrete-time ACM mediation models exhibited consistent mediation for all three pathways suggesting that 2-day lag in IDWG and preSBP responses can explain 2.11% and 4.41% of total effect association between temperature and mortality, respectively.

CONCLUSION

We corroborated the previously reported association between ambient temperature, ACHA and ACM. Our results foster the understanding of potential physiological linkages that may explain or suppress temperature-driven hospital admissions and mortality risks. Of note, concomitant changes in preSBP and IDWG may have little intermediary effect when analyzed in combined pathway models. These findings advance our assessment of candidate interventions to reduce the impact of outdoor temperature change on ESKD patients.

Time-varying risk associations of renin angiotensin system inhibitors with pneumonia and related deaths in a cohort of 252,616 patients with diabetes (2002-2019)

AIMS

To evaluate the time-varying and cumulative risk associations of renin-angiotensin-system-inhibitors (RASi) with pneumonia and related deaths in people with diabetes.

METHODS

This was a prospective analysis with propensity-score overlap-weighting of a territory-wide cohort (n = 252,616, 1.7 million person-years) and a register-based cohort (n = 13,017, 0.1 million person-years) of patients with diabetes in Hong Kong. We compared risk of pneumonia and related death in new-users of angiotensin-converting-enzyme-inhibitor (ACEi) and angiotensin-receptor-blocker (ARBs) with non-RASi users and new-users of calcium-channel-blockers as active comparator.

RESULTS

Amongst 252,616 people with diabetes (99.3% type 2 diabetes) in the population-based cohort with a mean follow-up of 6.7 years, 73,161 were new-ACEi-only users; 20,907 new-ARBs-only users; 38,778 ACEi/ARBs users; and 119,770 never-ACEi/ARBs. Time-varying RASi exposure was associated with reduced risk of pneumonia (HR = 0.78, 95% CI: 0.75-0.82) and pneumonia-related death (HR = 0.49, 0.46-0.53). The respective HRs for ARBs-only were 0.70 (0.62-0.78) and 0.41 (0.33-0.52) and that of ACEi-only were 0.98 (0.91-1.05) and 0.77 (0.68-86). The attenuated risk association of RASi use was time-invariant for pneumonia (P = 0.340) and time-varying for related-death (P < 0.001) with prevention of 0.6 (0.2-0.9) and 1.4 (1.0-1.6) per-1000-person-years events and deaths, respectively.

CONCLUSIONS

Long-term use of RASi, notably ARBs, was associated with reduced risk of pneumonia and related deaths in Chinese people with diabetes.

Emulating a target trial of proton pump inhibitors and dementia risk using claims data

BACKGROUND

Understanding the adverse effects of proton pump inhibitors (PPIs) is important due to their widespread use, but the available evidence for an increased dementia risk among patients taking PPIs is inconclusive. The present study aimed to estimate the causal effect of PPIs on the risk of dementia by target trial emulation and time-varying exposure modeling.

METHODS

Using claims data of 2,698,176 insured people of a large German statutory health insurer, we conceptualized a target trial in which individuals aged 40 years and older were classified as PPI initiators or non-initiators between 2008 and 2018, and followed until diagnosis of dementia, death, loss to follow-up or end of study. Incidence of dementia (ICD-10 codes F00, F01, F03, F05.1, G30, G31.0, G31.1, G31.9, and F02.8+G31.82) was defined applying a 1-year lag window. We used weighted Cox models to estimate the effect of PPI initiation vs. non-initiation on dementia risk and weighted pooled logistic regression to estimate the effect of time-varying use vs. non-use.

RESULTS

29,746 PPI initiators (4.4%) and 26,830 non-initiators (1.3%) were diagnosed with dementia. Comparing PPI initiation with no initiation, the hazard ratio (HR) for dementia was 1.54 (95% confidence interval [CI]: 1.51-1.58). The HR for time-dependent PPI use vs. non-use was 1.56 (95% CI: 1.50-1.63). Differentiated subtypes, including unspecified dementia, Alzheimer's disease (AD), and vascular dementia (VaD), showed increased risk by PPI initiation and time-varying PPI use.

CONCLUSIONS

This study suggests that PPI initiation and time-varying PPI use may increase overall dementia risk.

Associations of RBC and Serum Folate Concentrations with Cervical Intraepithelial Neoplasia and High-Risk Human Papillomavirus Genotypes in Female Chinese Adults

BACKGROUND

Although folate status is associated with cervical carcinogenesis, it is not clear whether folate deficiency is associated with risk of cervical intraepithelial neoplasia (CIN) progression and infection with high-risk human-papillomavirus (hrHPV).

OBJECTIVES

To evaluate the associations of RBC and serum folate concentrations with prevalence of CIN grades and hrHPV infection, their interactions with prevalence of CIN grades, and RBC folate with the risk of CIN1 progressing to CIN2.

METHODS

Using data from the Shanxi CIN cohort of 2304 female Chinese adults, we used logistic-regression model to estimate ORs and prevalence ratios (PRs) of RBC and serum folate concentrations with prevalence of CIN grades and hrHPV infection. Categoric and spline analyses were used to evaluate the dose-response relations. We estimated the association of RBC folate with risk of CIN1 progressing to CIN2 in the nested case-control cohort.

RESULTS

An inverse association was observed between increased RBC folate concentration and the odds of all CIN grades [quartile 1 (Q1) compared with Q4: OR: 2.28; 95% CI: 1.77, 2.93; Ptrend < 0.001]. Significant interaction of RBC folate and hrHPV infection was observed for prevalence of CIN2 or above (Pinteraction < 0.01). No associations were found between RBC and serum folate with PRs of hrHPV in each CIN grade. Over a median follow-up of 21.0 mo, RBC folate was associated with increased risk of CIN1 progressing to CIN2 (Q1 compared with Q4: OR: 3.86; 95% CI: 1.01, 14.76).

CONCLUSIONS

Our study indicates that RBC folate concentration is associated with prevalence of CIN grades and CIN1 progression in female Chinese adults. Maintenance of normal folate status is important for reducing the risk of CIN and its progression in women with or without hrHPV infection.