Pharmacoepidemiology and Drug Safety最新文献

Purpose: To identify the best method for selecting index dates when constructing external comparator arms (ECAs) from real-world data for comparison with single-arm trials (SATs).

Methods: We evaluated four approaches for index date selection-first eligible line, last eligible line, random eligible line, and all eligible lines-using causal inference reasoning, numerical examples and a simulation study. Simulations modeled survival across multiple lines of therapy under scenarios with varying eligibility patterns and treatment effects. Overall survival (OS) estimates comparing SAT and ECA populations were obtained using stratified Cox models and propensity-weighted Cox models, adjusted for line of therapy and patient state.

Results: Including all eligible lines produced unbiased OS estimates across scenarios. Selecting the last eligible line introduced substantial bias, while random selection led to moderate bias.

Conclusions: Using all eligible lines of therapy for each patient when constructing ECAs minimizes bias and preserves the SAT target population. Alternative methods can lead to biased estimates or, in the case of the first eligible line method, require changes to the clinical question that may shrink the SAT population. We recommend adopting the all eligible lines method with variance correction and adjustment for line of therapy to ensure valid comparative effectiveness analyses.

Objective: The aim of this study was to compare the efficacy and safety of monoclonal antibodies (mAb) for moderate-to-severe atopic dermatitis (AD).

Methods: The randomized controlled trials (RCTs) of monoclonal antibodies in the treatment of moderate-to-severe AD were searched in the database of PubMed, Embase, Web of Science and Cochrane Library, up to November 2024. The control group included placebo. The efficacy indicators were the percentage of patients achieving 50%, 75%, 90% improvement in Eczema Area and Severity Index score (EASI-50, EASI-75, EASI-90) and the percentage of patients with an Investigator Global Assessment (IGA) Score of 0 or 1 from baseline until the time of efficacy observation, and the percent change in Pruritus Numerical Rating Scale (NRS), EASI score, SCORing Atopic Dermatitis (SCORAD), and change in Percent Body Surface Area (BSA), Dermatology Life Quality Index (DLQI). The statistical analysis was performed by Stata14 and RevMan5.4. Data processing, network evidence plots, surface under the cumulative ranking curve (SUCRA) ranking, league plots and funnel plots were generated. Risk ratio (RR) and 95% confidence interval (95% CI) were used as effect sizes to analyze binary categorical variables.

Results: This study included 32 RCTs with 7588 patients. Spesolimab, Rademikibart, Dupilumab, Amlitelimab were more effective than placebo in EASI-50 (RRs ranging between 1.31 and 22.65), EASI-75(RRs ranging between 1.51 and 36.58), EASI-90(RRs ranging between 3.72 and 5.49), and the percentage of patients in IGA score of 0 or 1 (RRs ranging between 1.78 and 13.36). Dupilumab showed relatively good efficacy according to SUCRA ranking on percent change in NRS, EASI, SCORAD. Dupilumab exhibited a lower incidence of serious adverse events than placebo, and the difference was statistically significant (RR = 0.43, 95% CI 0.30-0.63). Other safety analysis results showed no statistical difference.

Conclusions: Through the analysis of the primary efficacy indicators, this network meta-analysis (NMA) study indicated that all monoclonal antibodies performed better than placebo. Based on the results of this study, Spesolimab, Rademikibart, Dupilumab, Amlitelimab were recommended treatment options with relatively good efficacy and safety.

Purpose: The ongoing global increase in opioid use necessitates studies examining long-term health impacts. Prospective cohorts frequently rely on self-reported medication use data which may be subject to several types of bias compared to more objective measurements. We evaluated the agreement between two opioid exposure measures in the UK Biobank (UKBB)-self-reported regular use and prescription-based indicators using linked general practitioner (GP) records.

Methods: Our analysis included 171 813 UKBB participants with linked prescription records. At baseline, participants reported medications taken regularly (weekly, monthly, every 3 months). We assessed agreement between self-reported regular opioid use and opioid prescription records prior to enrollment across various look-back periods and prescription counts. Logistic regressions assessed factors associated with omission and commission.

Results: Agreement was moderate to substantial between self-reported opioid use and prescription records. The strongest agreement was observed for ≥ 3 prescriptions in the past 365 days (Cohen's Kappa = 0.66). Subgroup analysis showed better agreement for analgesic opioids (Kappa = 0.43-0.63) than for non-analgesic opioids (Kappa = 0.25-0.34). Omission odds were highest with increasing months since the last record and lowest for individuals with chronic pain. Commission odds were highest for individuals with chronic pain and lowest in married/partnered individuals.

Conclusions: This analysis indicates that self-reported regular opioid use in the UKBB could be a valid indicator for identifying individuals with repeated prescriptions for analgesic opioids in the past year, while also capturing opioid use from non-linked sources. However, agreement was low for non-analgesic opioids, suggesting limited utility of self-report for capturing use of these medications.

Purpose: Anticholinergic medications and alcohol each independently increase the risk of community-acquired pneumonia (CAP). Whether non-anticholinergic neurocognitively active medications also increase risk, and if alcohol modifies these associations, remains unclear.

Methods: We conducted a nested case-control study using Veterans Aging Cohort Study (VACS)-National data. We identified 157 185 incident CAP cases requiring hospitalization between 2010 and 2022. Cases were matched 1:5 to controls without CAP on demographics, cohort entry date, and dwell time in the underlying cohort study using incidence density (risk-set) sampling. CAP index date was hospital admission for cases and the equivalent follow-up date for controls. Primary exposures were receipt of anticholinergic and non-anticholinergic neurocognitively active medications within 90 days prior to the index date. Concurrent alcohol use was based on self-reported measures in the year prior to the index date. We estimated odds ratios (ORs) for associations between medication use, alcohol consumption, and CAP using logistic regression, adjusting for confounders.

Results: Median age was 69 years (interquartile range 62-78); 97% were male. Both medication types were independently associated with increased odds of CAP (anticholinergic: OR 1.62, 95% CI 1.57-1.67; non-anticholinergic: OR 1.61, 95% CI 1.57-1.66). Concurrent alcohol use modified these associations. For anticholinergics, ORs were 1.74 (95% CI 1.66-1.83) for at-risk consumption and 2.13 (95% CI 1.96-2.31) for hazardous/binge consumption. For non-anticholinergics, ORs were 1.74 (95% CI 1.67-1.81) and 2.20 (95% CI 2.06-2.34), respectively.

Conclusions: Non-anticholinergic neurocognitively active medications showed similar CAP association patterns as anticholinergics, with the highest odds among those consuming alcohol. These findings highlight the need for caution when prescribing these medications and incorporating alcohol use into risk-benefit assessments.

Purpose: Vulvovaginal candidiasis is highly prevalent among pregnant women, and its treatment is crucial. However, there is a lack of evidence regarding the risks on fetal outcomes associated with the use of antifungal medications during the first trimester of pregnancy in Japan. This study examined the association between topical antifungal use during the first trimester of pregnancy and the risk of major congenital malformations (MCMs) in infants using a Japanese database.

Methods: We conducted a cohort study using a pregnancy cohort nested in the JMDC Claims Database from Japan. This dataset included 12 472 women who gave birth between 2010 and 2019 and were diagnosed with vulvovaginal candidiasis. Among the antifungal medications frequently dispensed or prescribed during the first trimester of pregnancy in this cohort, miconazole, oxiconazole, and isoconazole were assessed for the risk of MCMs, using clotrimazole, an antifungal medication with established safety during pregnancy, as a reference.

Results: Among 12 472 women, the overall prevalence of MCMs was 5.8% (n = 249) in women exposed to topical antifungals, while unexposed were 6.2% (n = 508). Using propensity score overlap weight (wOR), no increased risk of MCMs in infants was observed in pregnancies exposed to oxiconazole, isoconazole, and miconazole compared to clotrimazole (overlap weighted odds ratio [95% confidence interval]: 0.875 [0.599-1.277], 1.001 [0.611-1.640], and 0.887 [0.497-1.581], respectively).

Conclusion: There was no significant association between topical antifungal use during the first trimester of pregnancy and the risk of MCMs in infants in Japan.

Objective: Randomized clinical trials show that subcutaneous semaglutide is modestly superior to dulaglutide in reducing HbA1c and body weight, but no trial has compared their effectiveness on hard cardiovascular outcomes. This study aimed to examine whether semaglutide and dulaglutide differ in cardiovascular effectiveness.

Research design and methods: This new-user, active-comparator cohort study used nationwide population-based Danish healthcare data to emulate a target trial of adults with type 2 diabetes receiving standard care who initiated subcutaneous semaglutide compared with dulaglutide. Up to five semaglutide initiators were matched to one dulaglutide initiator on a propensity score estimated from 52 variables. The outcome was a major adverse cardiovascular event (MACE), including myocardial infarction, ischemic stroke, heart failure, coronary revascularization, and cardiovascular death. In per-protocol analyses, Aalen-Johansen estimates were used to calculate risks, risk differences, and risk ratios at 3 years, accounting for informative censoring at nonadherence to the assigned treatment via time-varying inverse probability of censoring weights.

Results: The semaglutide group included 2535 individuals, and the dulaglutide group 569 (median age [IQR], 61 [52-71] years; 1105 female individuals [36%]). Within 3 years, the risk of MACE was 6.0% (95% CI, 4.5%-7.8%) in the semaglutide group and 6.2% (95% CI, 4.0%-8.9%) in the dulaglutide group, corresponding to a risk difference of -0.2% (95% CI, -3.2% to 2.8%) and a risk ratio of 0.97 (95% CI, 0.59-1.61).

Conclusions: This target trial emulation did not provide evidence for a substantial difference in cardiovascular outcomes between individuals with type 2 diabetes initiating semaglutide and dulaglutide.

Purpose: Evidence on head-to-head comparison between direct oral anticoagulants (DOACs) is lacking, and the reasons for choosing one type of DOAC and switching from one DOAC to another are scarce. This study investigated the use of DOACs in an unselected population in Northern Italy during a recent period.

Methods: Using the health administrative database of the Lombardy region, subjects aged 45 years and older who started DOAC therapy between 2019 and 2022 were included in the analysis. Logistic regression analysis was used to evaluate predictors associated with DOAC prescription, and results were presented as ORs with 95% CI. DOAC switching was assessed by estimating the prevalence and cumulative incidence according to the first prescribed DOAC.

Results: Overall, 159 993 new users for DOAC were identified. Apixaban users increased from 29.0% to 34.3%, whereas dabigatran users decreased from 20.7% to 11.5% over time. Across all pair-wise comparisons, older age and female sex were predictors for edoxaban prescription. Comorbidities were mainly associated with the use of apixaban; however, dabigatran was preferred in patients with a history of ischemic heart disease or myocardial infarction and rivaroxaban in those with peripheral artery disease. Both apixaban and edoxaban were preferentially prescribed to patients with a history of bleeding. The switching rate of DOACs was 5.6% with apixaban as the most preferred drug as a second choice. Dabigatran was mainly chosen as a second DOAC after a vascular ischaemic event.

Conclusion: Given the lack of evidence on factors influencing clinician behavior in the use of DOACs, our findings provide insight into this topic in a real-world setting. As the use of these agents increases, further evidence is needed to better explore this issue. Our data could contribute to the development of recommendations in clinical practice.

Introduction: The COVID-19 pandemic caused unprecedented disruptions in healthcare delivery, and changes in medication utilization patterns. While previous studies examined specific therapeutic classes or populations, there is limited longitudinal evidence on medication trends among young adults throughout and beyond the pandemic.

Aim: To analyze trends in medication dispensation before, during, and after the COVID-19 pandemic among young adults.

Methods: We conducted a population-based, retrospective cohort study including active-duty Israeli Defense Forces personnel between January 2017 and August 2023. Monthly dispensing rates per 1000 persons were analyzed using an interrupted time series (ITS) design, implemented via generalized linear models with log link and population offsets. Models included linear time trends, month fixed effects to account for seasonality, and negative binomial fallback for overdispersion. Pre-pandemic data (January 2017-March 2020) were used to estimate baseline trends, from which counterfactual predictions were generated for March 2020-August 2022. Goodness-of-fit was evaluated with RMSE and MAPE.

Results: Pre-pandemic trends varied across therapeutic groups. Adrenergic inhalants (IRR 1.008, 95% CI 1.004-1.011, p = 0.0001), antidiarrheals (IRR 1.005, 95% CI 1.001-1.008, p = 0.004), and ADHD medications (IRR 1.024, 95% CI 1.020-1.027, p < 0.001) exhibited significant upward slopes, whereas antibacterials, antidepressants and hormonal contraceptives showed no significant baseline trend. Seasonality was significant for all groups (p < 0.001). During the pandemic, cumulative differences revealed excesses for adrenergic inhalants (+93.98 per 1000), antidepressants (+87.03), and hormonal contraceptives (+679.21), alongside deficits for antibacterials (-201.99), antidiarrheals (-112.89), and ADHD medications (-294.69).

Conclusions: Medication usage patterns can be classified into three classes: medications affected by the pandemic due to the inciting pathogen, disease symptoms, or pandemic social disruption; medications unaffected by the pandemic, affected by global disease trends; and medications with a trend change whose relation to the pandemic is unclear. These findings offer a novel framework for anticipating and managing medication needs in future pandemics.

Background: In pharmacoepidemiologic studies of COVID-19, there were concerns about bias from residual confounding. We investigated the effects of inhaled corticosteroids (ICS) on COVID-19 outcomes, applying high-dimensional propensity scores (HDPS) to adjust for unmeasured confounding.

Methods: We selected patients with chronic obstructive pulmonary disease on 01 March 2020 from Clinical Practice Research Datalink (CPRD) Aurum, comparing ICS/LABA/(+/-LAMA) and LABA/LAMA users. ICS effects on the outcomes COVID-19 hospitalisation and death were assessed through IPT-weighted and unweighted Cox regression. HDPS were estimated from primary care observations, prescriptions and hospitalisations. SNOMED-CT codes and dictionary of medicines and devices codes from CPRD Aurum were mapped to International Classification of Disease 10th revision codes and British National Formulary paragraphs, respectively. We estimated propensity scores (PS) combining prespecified and HDPS covariates, selecting the top 100, 250, 500, 750 and 1000 covariates ranked by confounding potential.

Results: When excluding triple therapy users, conventional PS-weighted estimates showed weak evidence of increased COVID-19 hospitalisation risk among ICS users (HR 1.19 [95% CI: 0.92-1.54]). Results varied slightly based on the number of covariates included in HDPS (HR using 100 HDPS covariates excluding triple therapy 1.01 [95% CI: 0.76-1.33], HR using 250 HDPS covariates excluding triple therapy 1.24 [95% CI: 0.83-1.87]). Conventional PS-weighted models showed weak evidence of a harmful association of ICS with COVID-19 death when excluding triple therapy users (HR 1.24 [95% CI: 0.87-1.75]). HDPS-weighting moved estimates toward the null (HR using 250 HDPS covariates excluding triple therapy 1.08 [95% CI: 0.73-1.59]).

Conclusions: HDPS may have better controlled confounding for COVID-19 deaths in this case. HDPS results can be sensitive to the number of covariates included, highlighting the importance of sensitivity analyses.

Background: There is little evidence about signal detection using UK primary care electronic health records (EHRs). The self controlled case series (SCCS) is one of the most promising methods for drug safety signal detection using real world data, and incorporating active comparators could potentially improve its performance by addressing confounding by indication.

Objectives: This study aims to evaluate the performance of the SCCS with and without active comparators for signal detection using the UK Clinical Practice Research Datalink (CPRD) Aurum.

Methods: We applied the SCCS to macrolide and fluoroquinolone antibiotics, using amoxicillin and cefalexin as active comparators. In total seven drugs, and 30 outcomes from all organ classes were selected. We developed a reference set of 104 positive controls and 58 negative controls, using a taxonomy framework to ensure the selected drug outcome pairs are theoretically well suited to the SCCS design. Two-year observation periods with a 30-day risk window after each dispensing were used. Diagnostic performance was measured using sensitivity and specificity with respect to the product labels.

Results: The sensitivity and specificity of the SCCS without active comparator in the 2017/2018 observation period were 0.57 and 0.77 when limited to pairs with satisfactory power. Specificity increased up to 0.89 with active comparators, however sensitivity decreased to 0.18. Five drug-outcome pairs were signals of disproportionality before they were present on labels.

Conclusions: Using a carefully designed reference set of drug-outcome pairs well suited to the study design, the SCCS performed moderately well for signal detection in CPRD. Whilst active comparators effectively reduced confounding by indication, they also reduced the number of correctly identified positive controls, due to a reduction in power. We showed some evidence that SCCS is able to highlight SDRs before they were present on labels.