Category: Statistics

Arch Pathol Lab Med. 2025 Apr 28. doi: 10.5858/arpa.2024-0448-RA. Online ahead of print.

ABSTRACT

CONTEXT.—: Cervical squamous neoplasia runs the gamut from low-risk intraepithelial processes to aggressive invasive malignancies. A variety of biomarkers can be enlisted to help diagnose, prognosticate, and inform treatment of these lesions. There are ongoing controversies about diagnostic and prognostic biomarker use in squamous intraepithelial lesions, and many pathologists are new to predictive biomarker interpretation in invasive cervical lesions.

OBJECTIVE.—: To provide practical guidance on the appropriate use of diagnostic, prognostic, and predictive biomarkers in cervical squamous intraepithelial lesions and invasive carcinomas.

DATA SOURCES.—: Peer-reviewed literature and the author’s personal experience.

CONCLUSIONS.—: Diagnostic biomarkers such as p16 and human papillomavirus E6/E7 messenger RNA in situ hybridization can have value in the diagnosis of squamous intraepithelial neoplasia, but there are important caveats to their use and interpretation. No prognostic biomarkers have yet demonstrated statistically durable significance for risk stratification of low-grade squamous intraepithelial lesions. Programmed death ligand-1 immunohistochemistry and tumor mutational burden testing are US Food & Drug Administration-approved predictive biomarkers that can be enlisted for the identification of invasive cervical squamous carcinomas that may respond to checkpoint inhibitor-based immunotherapy, whereas human epidermal growth factor receptor 2 (HER2) immunohistochemistry can identify optimal candidates for conjugated anti-HER2 therapies.

PMID:40289713 | DOI:10.5858/arpa.2024-0448-RA

Pharm Stat. 2025 May-Jun;24(3):e70015. doi: 10.1002/pst.70015.

ABSTRACT

The term “minimal clinically important difference” (MCID), though defined as the smallest change in an outcome that is meaningful to the patient, is often used to interpret differences between treatment groups. It is in this context that the limitations of MCID are discussed, which include: the omission of the role of time in its definition for progressive diseases; the unsuitability of adopting MCID derived from open-label studies for randomized, placebo-controlled, blinded studies; the unreliability of MCID in rare disease trials; challenges in interpretation when placebo patients also achieve MCID; the failure to account for how differences in patient populations affect MCID (e.g., inclusion or exclusion of patients on prior treatment); not recognizing the connection between the true treatment effect, MCID and power; lack of consideration of differences in analysis methods (e.g., the extent of missing data and how it is handled); and the limitations of an MCID-based responder analysis. Therefore, the recommendation made is to prospectively define a customized MCID that addresses each deficit. If the deficits cannot be adequately resolved, then the recommendation is that trial results should be interpreted without reference to MCID.

PMID:40289700 | DOI:10.1002/pst.70015

Influenza Other Respir Viruses. 2025 May;19(5):e70097. doi: 10.1111/irv.70097.

ABSTRACT

BACKGROUND: RSV incidence in adults is frequently underestimated due to non-specific symptomatology, limited standard-of-care testing, and lower test sensitivity compared to infants. We conducted a retrospective observational study to estimate RSV-attributable incidence of specific types of cardiorespiratory hospitalizations among adults in Germany between 2015 and 2019.

METHODS: Information on hospitalizations and the number of people at risk of hospitalization (denominator) was gathered from a Statutory Health Insurance database. A quasi-Poisson regression model accounting for periodic and aperiodic time trends and virus activity was fitted to estimate the RSV-attributable incidence rate (IR) of four specific cardiovascular hospitalizations (arrhythmia, ischemic heart diseases, chronic heart failure exacerbations, and cerebrovascular diseases) and four specific respiratory hospitalizations (influenza/pneumonia, bronchitis/bronchiolitis, chronic lower respiratory tract diseases, and upper respiratory tract diseases).

RESULTS: The estimated RSV-attributable IRs of hospitalizations generally increased with age. Among estimated cardiovascular hospitalizations in adults aged ≥ 60 years, arrhythmia and ischemic heart diseases accounted for the highest incidence of RSV-attributable events, followed by chronic heart failure exacerbation, with annual IR ranges of 157-260, 133-214, and 105-169 per 100,000 person-years, respectively. The most frequent RSV-attributable respiratory hospitalizations in adults aged ≥ 60 years were estimated for chronic lower respiratory tract diseases and bronchitis/bronchiolitis, with annual IR ranges of 103-168 and 77-122 per 100,000 person-years, respectively.

CONCLUSIONS: RSV causes a considerable burden of respiratory and cardiovascular hospitalizations in adults in Germany, similar to other respiratory viruses (e.g., influenza and SARS-CoV-2). This highlights the need to implement effective prevention strategies, especially for older adults.

PMID:40289699 | DOI:10.1111/irv.70097

Ann Med. 2025 Dec;57(1):2496798. doi: 10.1080/07853890.2025.2496798. Epub 2025 Apr 28.

ABSTRACT

BACKGROUND: Psychomotor impairments due to alcohol consumption may lead to a series of negative consequences. However, the influence of sex and ALDH2 polymorphism on psychomotor dysfunction has not yet been investigated.

METHODS: One-hundred and three participants, genotyped for ALDH2 rs671, were administered a dose of 1.0 g/kg of white spirits. The blood ethanol concentration (BEC) and acetaldehyde concentration (BAAC) were measured at specific time intervals before and after alcohol consumption. Additionally, auditory simple reaction time (ASRT), visual choice reaction time (VCRT), pursuit tracking task (PTT) and digit-symbol substitution test (DSST) were used to evaluate psychomotor function. Linear mixed-effects model was used to analyze the effects of sex and the ALDH2 genotype on alcohol metabolism and psychomotor function..

RESULTS: Acetaldehyde metabolism depended on both ALDH2 genotype and sex. Women with ALDH2*1/*1 genotype exhibited 2.21 to 18.27 µmol/L higher BAAC levels than men with the same genotype. Conversely, among participants with ALDH2*1/*2 genotype, BAAC levels of women were 0.25 to 31.32 µmol/L lower than men. The impact of ALDH2 genotype on psychomotor function varied across the four tests. VCRT increased significantly in men with ALDH2*1/*2 genotype compared to those with ALDH2*1/*1 at 2-4 h post-consumption. In the PTT test, the percentage of time on target decreased by 3.83% and 3.11% in women relative to men at 1 and 2 h post-consumption, respectively. Notably, ASRT performance was significantly correlated with BAAC levels. No effects of ALDH2 genotype and sex were observed on DSST performance.

CONCLUSIONS: ALDH2 genotype and sex independently or interactively contribute to alcohol-related psychomotor impairment.

PMID:40289679 | DOI:10.1080/07853890.2025.2496798

Addiction. 2025 Apr 28. doi: 10.1111/add.70080. Online ahead of print.

ABSTRACT

AIMS: We measured the effects of public health-oriented cannabis access compared with the illegal market on cannabis use and related mental health outcomes in adult cannabis users.

DESIGN: This was a two-arm, parallel group, open-label, randomized controlled trial. Follow-up outcome measurement took place after 6 months.

SETTING: The study was conducted in Basel-Stadt, Switzerland.

PARTICIPANTS: A total of 378 adult (aged ≥18 years) cannabis users were enrolled and randomized between August 2022 and March 2023, although only 374 users who completed baseline measures could be included.

INTERVENTION AND COMPARATOR: Participants were randomly assigned to the intervention group with public health-oriented recreational cannabis access in pharmacies (regulated cannabis products, safer use information, voluntary counseling, no advertisement; 189/188) or the illegal market control group (continued illicit cannabis sourcing; 189/186).

MEASUREMENTS: The primary outcome was self-reported severity of cannabis misuse after 6 months, as measured by the Cannabis Use Disorders Identification Test – Revised (range 0-32). Secondary outcomes involved depressive, anxiety, and psychotic symptoms, cannabis consumption amount, alcohol, and drug use.

FINDINGS: Ten participants were not followed (2.7%). Primary analysis included those with complete data (182 vs. 182). There was some evidence of a difference in cannabis misuse between the legal cannabis intervention group (mean [M] = 10.1) and the illegal market control group (M = 10.9; β = -0.69, 95% confidence interval [CI] = -1.4 to 0.0, P = 0.052). These results were supported by an intention-to-treat multiple imputation analysis (n = 374). Additional sub-group analysis by whether the participant used other drugs or not suggested that any reduction in cannabis misuse was confined to those in the legal cannabis intervention group who used other drugs (P_Interaction < 0.001). We found no statistically significant changes in any of the secondary outcomes.

CONCLUSIONS: Public health-oriented recreational cannabis access may decrease cannabis use and cannabis-related harms, especially among those using other drugs.

PMID:40289676 | DOI:10.1111/add.70080

Twin Res Hum Genet. 2025 Apr 28:1-8. doi: 10.1017/thg.2025.15. Online ahead of print.

ABSTRACT

Incorporating genetic data from diverse populations is crucial for understanding genetic contributions to diseases and ensuring health equity in healthcare practices. However, existing reference panels either capture a limited number of populations or have small sample sizes. We examine the UK Biobank’s performance as a reference for clustering genetically similar individuals. Leveraging data from participants of diverse origins, we aim to improve population representation and mitigate bias caused by the limited number of populations in other reference panels. We combined countries of birth and ethnic backgrounds data fields from the UK Biobank and genetic information to infer genetically similar population labels. A random forest model was then trained on genetic principal components to identify each individual’s most genetically similar population. The model’s performance was validated using the 1000 Genomes and the CARTaGENE biobank data. We identified more diverse reference populations than present in datasets such as 1000 Genomes, covering 19 populations worldwide. Our model achieved medium to high precision and recall for most labeled populations, although lower rates were observed in closely related groups. For instance, we identified 519 people in CARTaGENE most genetically similar to the Middle Eastern reference sample derived in the UK Biobank (there are no Middle Eastern samples in 1000 Genomes), yielding an 81.1% precision and a 97.0% recall rate compared to demographic-based information. This practical approach of clustering genetically similar individuals utilizing existing biobank data may facilitate downstream analyses, such as genomewide association studies or polygenic risk scores in underrepresented populations in genetic studies.

PMID:40289653 | DOI:10.1017/thg.2025.15

Psychol Med. 2025 Apr 28;55:e119. doi: 10.1017/S0033291725000947.

ABSTRACT

BACKGROUND: This study examines the prospective associations of alcohol and drug misuse with suicidal behaviors among service members who have left active duty. We also evaluate potential moderating effects of other risk factors and whether substance misuse signals increased risk of transitioning from thinking about to attempting suicide.

METHOD: US Army veterans and deactivated reservists (N = 6,811) completed surveys in 2016-2018 (T1) and 2018-2019 (T2). Weights-adjusted logistic regression was used to estimate the associations of binge drinking, smoking/vaping, cannabis use, prescription drug abuse, illicit drug use, alcohol use disorder (AUD), and drug use disorder (DUD) at T1 with suicide ideation, plan, and attempt at T2. Interaction models tested for moderation of these associations by sex, depression, and recency of separation/deactivation. Suicide attempt models were also fit in the subgroup with ideation at T1 (n = 1,527).

RESULTS: In models controlling for socio-demographic characteristics and prior suicidality, binge drinking, cannabis use, prescription drug abuse, illicit drug use, and AUD were associated with subsequent suicidal ideation (AORs = 1.42-2.60, ps < .01). Binge drinking, AUD, and DUD were associated with subsequent suicide plan (AORs = 1.23-1.95, ps < .05). None of the substance use variables had a main effect on suicide attempt; however, interaction models suggested certain types of drug use predicted attempts among those without depression. Additionally, the effects of smoking/vaping and AUD differed by sex. Substance misuse did not predict the transition from ideation to attempt.

CONCLUSIONS: Alcohol and drug misuse are associated with subsequent suicidal behaviors in this population. Awareness of differences across sex and depression status may inform suicide risk assessment.

PMID:40289652 | DOI:10.1017/S0033291725000947

Psychol Med. 2025 Apr 28;55:e122. doi: 10.1017/S003329172500090X.

ABSTRACT

BACKGROUND: Individuals with long-term physical health conditions (LTCs) experience higher rates of depression and anxiety. Conventional self-report measures do not distinguish distress related to LTCs from primary mental health disorders. This difference is important as treatment protocols differ. We developed a transdiagnostic self-report measure of illness-related distress, applicable across LTCs.

METHODS: The new Illness-Related Distress (IRD) scale was developed through thematic coding of interviews, systematic literature search, think-aloud interviews with patients and healthcare providers, and expert-consensus meetings. An internet sample (n = 1,398) of UK-based individuals with LTCs completed the IRD scale for psychometric analysis. We randomly split the sample (1:1) to conduct: (1) an exploratory factor analysis (EFA; n = 698) for item reduction, and (2) iterative confirmatory factor analysis (CFA; n = 700) and exploratory structural equation modeling (ESEM). Here, further item reduction took place to generate a final version. Measurement invariance, internal consistency, convergent, test-retest reliability, and clinical cut-points were assessed.

RESULTS: EFA suggested a 2-factor structure for the IRD scale, subsequently confirmed by iteratively comparing unidimensional, lower order, and bifactor CFAs and ESEMs. A lower-order correlated 2-factor CFA model (two 7-item subscales: intrapersonal distress and interpersonal distress) was favored and was structurally invariant for gender. Subscales demonstrated excellent internal consistency, very good test-retest reliability, and good convergent validity. Clinical cut points were identified (intrapersonal = 15, interpersonal = 12).

CONCLUSION: The IRD scale is the first measure that captures transdiagnostic distress. It may aid assessment within clinical practice and research related to psychological adjustment and distress in LTCs.

PMID:40289643 | DOI:10.1017/S003329172500090X

Int J Low Extrem Wounds. 2025 Apr 28:15347346251337264. doi: 10.1177/15347346251337264. Online ahead of print.

ABSTRACT

The current study aimed to evaluate the clinical outcomes of patients admitted for diabetic foot ulcers (DFUs) located in the heel. The study is a retrospective observational study including a population of patients admitted for DFUs from April 2024 to September 2024. According to the wound location, patients were divided in two groups: those with heel ulcers and those with forefoot/midfoot ulcers without calcaneal involvement. All patients have been managed through a limb salvage protocol in the respect of international guidelines. The following hospital outcomes were evaluated: minor and major amputation, mortality, length of stay, need for regenerative surgery by using dermal-epidermal graft. Overall, 150 patients were included. The mean age was 70.2 ± 12.2 years, most patients were male (76.0%), had type 2 diabetes (92.7%) with a mean duration of 22.1±13.2 years; 27 (18%) patients had heel ulcers, while 123 (82%) had forefoot/midfoot lesions. Outcomes for patients with heel location and without were: minor amputation (18.5 vs 32.5%, p = 0.1), major amputation (7.4 vs 0.8%, p = 0.02), mortality (0 vs 0.8%, p = 0.5), length of hospital stay (17.5 ± 8.5 vs 14.4 ± 8.7 days, p = 0.08), need for regenerative therapy using dermal-epidermal substitutes (48.1 vs 19.5%, p = 0.003) respectively. In addition, heel ulcer was found to be an independent predictor for major amputation [OR 5.06, CI95% (3.1-11.4), p = 0.02] and length of stay [OR 6 CI95% (3.6-10.9), p = 0.003]. In patients admitted for DFUs, wounds located in the heel were associated to an increased risk of major amputation, need for regenerative therapy and length of stay (even though in the limit of statistical difference) than wounds not located in the heel. These data underline the need for tailored management strategies in this high-risk subgroup of patients.

PMID:40289631 | DOI:10.1177/15347346251337264

Stat Methods Med Res. 2025 Apr 28:9622802251314195. doi: 10.1177/09622802251314195. Online ahead of print.

ABSTRACT

Clinical trials often face the challenge of post-randomization events, such as the initiation of rescue therapy or the premature discontinuation of randomized treatment. Such events, called “intercurrent events” (ICEs) in ICH E9(R1), may influence the estimation and interpretation of treatment effects. According to ICH E9(R1), there are five strategies for handling ICEs. This study focuses on the composite strategy, which incorporates ICEs in the outcome of interest and defines the treatment effects using composite endpoints that combine the measured continuous variables and ICEs. An advantage of this strategy is that it avoids the occurrence of missing data because they are defined as part of the outcome of interest. In this study, we propose a new composite estimand: the difference in the partial areas under the curves (pAUCs) of the cumulative distribution function. While the pAUC is closely related to the trimmed mean approach proposed by Permutt and Li, it offers the advantage of allowing pre-specification of the cutoff value for a “good” response based on clinical considerations. This ensures that the pAUC can be calculated irrespective of the proportion of ICEs. We describe the causal interpretation of our method and its relationship with two other strategies (treatment policy and hypothetical strategies) using a potential outcome framework. We present simulation results in which our method performs reasonably well compared to several existing approaches in terms of type I error, power, and the proportion of undefined test statistics.

PMID:40289630 | DOI:10.1177/09622802251314195