Category: Statistics

Stat Med. 2023 Aug 1. doi: 10.1002/sim.9853. Online ahead of print.

ABSTRACT

In modern medicine, medical tests are used for various purposes including diagnosis, disease screening, prognosis, and risk prediction. To quantify the performance of the binary medical test, we often use sensitivity, specificity, and negative and positive predictive values as measures. Additionally, the $F_1$ -score, which is defined as the harmonic mean of precision (positive predictive value) and recall (sensitivity), has come to be used in the medical field due to its favorable characteristics. The $F_1$ -score has been extended for multi-class classification, and two types of $F_1$ -scores have been proposed for multi-class classification: a micro-averaged $F_1$ -score and a macro-averaged $F_1$ -score. The micro-averaged $F_1$ -score pools per-sample classifications across classes and then calculates the overall $F_1$ -score, whereas the macro-averaged $F_1$ -score computes an arithmetic mean of the $F_1$ -scores for each class. Additionally, Sokolova and Lapalme ${}^1$ gave an alternative definition of the macro-averaged $F_1$ -score as the harmonic mean of the arithmetic means of the precision and recall over classes. Although some statistical methods of inference for binary and multi-class $F_1$ -scores have been proposed, the methodology development of hypothesis testing procedure for them has not been fully progressing yet. Therefore, we aim to develop hypothesis testing procedure for comparing two $F_1$ -scores in paired study design based on the large sample multivariate central limit theorem.

PMID:37527903 | DOI:10.1002/sim.9853

BMJ Open. 2023 Aug 1;13(8):e069814. doi: 10.1136/bmjopen-2022-069814.

ABSTRACT

OBJECTIVES: Multimorbidity challenges healthcare systems. In Germany, coordination of healthcare for older multimorbid patients remains unstructured.This study aims to identify key themes in the healthcare of these patients and the inter-relationships between them.

DESIGN: Framework analysis of six cases based on 1-year data of primary and secondary care, patient-answered questionnaires and video material.

SETTING: Southern Germany.

PARTICIPANTS: Six multimorbid older patients participating in a randomised controlled trial that compared usual care with a local, collaborative, stepped and personalised care management approach for older people with chronic diseases (LoChro-trial). The LoChro care intervention involved a care manager who assisted participants in self-management. The primary outcome was a composite of functional health and depressive symptoms at 12 and 18 months. The LoChro-intervention had no effect on the primary outcome.

PRIMARY OUTCOME MEASURE: Key themes in the healthcare of older patients with multimorbidity and the inter-relationships between them.

METHODS: One-year data included diagnoses, treatment plans, examinations, assessments and discharge reports. Patient perspectives were assessed using the Patient Assessment of Chronic Illness Care. In three cases, videos of the LoChro intervention showed patients describing their health needs. These data were evaluated by three doctors and public health researchers. Using framework analysis, recurring themes influencing the healthcare situation of multimorbid older patients and their inter-relationships were identified.

RESULTS: Participants had an average age of 77, with 13 diagnoses, taking eight medications regularly. The five key themes describing the healthcare situation of these multimorbid patients were as follows: insufficient coordination, overuse and underuse of medical care, doctor and patient roles. Each theme covered three to four subcategories. The most significant inter-relationships between these themes were a lack of coordination leading to overuse and underuse of medical care. These were characterised by redundant inpatient stays, potential prescribing omissions and missed examinations. Deficiencies in vaccinations and secondary prevention were also demonstrated.

CONCLUSION: Coordination of care for multimorbid older patients in Germany is still deficient. Future healthcare arrangements should be explored with the participation of physicians and patients.

TRIAL REGISTRATION NUMBER: LoChro trial: DRKS00013904.

PMID:37527899 | DOI:10.1136/bmjopen-2022-069814

BMJ Open. 2023 Aug 1;13(8):e071318. doi: 10.1136/bmjopen-2022-071318.

ABSTRACT

INTRODUCTION: Chronic kidney disease (CKD) affects 30 million Americans. Early management focused on blood pressure (BP) control decreases cardiovascular morbidity and mortality. Less than 40% of patients with CKD achieve recommended BP targets due to many barriers. These barriers include a lack of understanding of the implications of their diagnosis and how to optimise their health.This cluster randomised control trial hypothesises that the combination of early primary care CKD education, and motivational interviewing (MI)-based health coach support, will improve patient behaviours aligned with BP control by increasing patient knowledge, self-efficacy and motivation. The results will aid in sustainable interventions for future patient-centric education and coaching support to improve quality and outcomes in patients with CKD stages 3-5. Outcomes in patients with CKD stages 3-5 receiving the intervention will be compared with similar patients within a control group. Continuous quality improvement (CQI) and systems methodologies will be used to optimise resource neutrality and leverage existing technology to support implementation and future dissemination. The innovative approach of this research focuses on the importance of a multidisciplinary team, including off-site patient coaching, that can intervene early in the CKD care continuum by supporting patients with education and coaching.

METHODS AND ANALYSIS: We will test impact of BP control when clinician-delivered education is followed by 12 months of MI-based health coaching. We will compare outcomes in 350 patients with CKD stages 3-5 between intervention and control groups in primary care. CQI and systems methodologies will optimise education and coaching for future implementation and dissemination.

ETHICS AND DISSEMINATION: This study was approved by the University of Michigan Institutional Review Boards (IRBMED) HUM00136011, HUM00150672 and SITE00000092 and the results of the study will be published on ClinicalTrials.gov, in peer-reviewed journals, as well as conference abstracts, posters and presentations.

TRIAL REGISTRATION NUMBER: NCT04087798.

PMID:37527897 | DOI:10.1136/bmjopen-2022-071318

BMJ Open. 2023 Aug 1;13(8):e067860. doi: 10.1136/bmjopen-2022-067860.

ABSTRACT

OBJECTIVES: Foster children are disproportionately exposed to complex trauma, which may lead to multifaceted impairments that manifest in comorbid emotional and behavioural problems. As little is known about the interactions between comorbid disorders over time, the present study aims to explore the co-development of traumatic stress (TS) symptoms and externalising behaviour problems (EBP), as well as the influence of complex trauma operationalised as cumulative child maltreatment (CM).

SETTING: As part of a 3-year longitudinal study, children from six foster care facilities in Lower Austria were interviewed at three measurement points.

PARTICIPANTS: Of, in total, 263 participating children, the data of 124 children aged 10-18 years (M=13.5, 28% female) could be analysed.

PRIMARY AND SECONDARY OUTCOME MEASURES: Latent growth curve models were used to examine the co-development of TS symptoms (International Trauma Questionnaire) and EBP (Child Behaviour Checklist) over time; gender, age and cumulative CM (Childhood Trauma Questionnaire) acted as time-invariant covariates.

RESULTS: While average TS symptoms decreased over time, EBP remained stable. Findings revealed that the initial severity of EBP was both related to the initial severity of TS symptoms and predictive of their rate of change. Cumulative CM was a significant predictor of initial TS symptoms and EBP even after controlling for age and gender, but not for the rates of change.

CONCLUSIONS: Taken together, our results indicate that EBP and TS symptoms are not only cross-sectionally associated but interact with each other over time. Furthermore, an underlying complex trauma could at least partly determine the severity of the two symptom groups. In accordance with a trauma-informed care approach, our study highlights the importance of trauma-specific screening of high-risk children with complex or diffuse symptoms and argues for the benefits of treatments that focus on improving emotion regulation and social skills in addition to addressing trauma.

PMID:37527893 | DOI:10.1136/bmjopen-2022-067860

BMJ Open. 2023 Aug 1;13(8):e070913. doi: 10.1136/bmjopen-2022-070913.

ABSTRACT

OBJECTIVES: This pilot study assessed whether a peer-supported, WhatsApp-assisted lifestyle modification intervention for weight reduction is feasible to execute a definitive trial.

DESIGN: A mixed-methods, single group, pretest and post-test, quasi-experimental study.

SETTING: Azam Basti, an urban slum in Karachi, Pakistan.

PARTICIPANTS: Fifty participants (males and females aged 20-60) with a body mass index of >23 kg/m², along with their nominated peers from the same family.

INTERVENTION: Using motivational interviewing techniques, a trained nutritionist delivered the lifestyle modification intervention to the participants and peers for 3 days after the baseline assessment and then once monthly for 1 year. The intervention was delivered in groups using WhatsApp voice calls. The education sessions mainly focused on dietary modifications, physical activity advice and peer-support assignments to achieve a 5% wt loss from the participant’s initial body weight.

OUTCOMES: The feasibility measures included screening, recruitment, retention and monthly interview response rates. At 1 year, in-depth interviews (IDIs) with participants and peers were conducted to explore the facilitators, barriers, acceptability and experiences of the intervention. Changes in weight, calorie intake/day and calorie expenditure/day were also assessed.

RESULTS: The recruitment and retention rates were 32% (n=50/156) and 78% (n=39/50), respectively, while the response rate for monthly interviews ranged between 66% (n=33) and 94% (n=47). The mean weight loss at 1 year was 2.2 kg, and the reduction in mean calorie intake was 386 kcal/day. There were no changes in the mean calorie expenditure. During the IDIs, participants and peers reported intervention via WhatsApp and peer support as convenient, flexible and supportive.

CONCLUSIONS: The quantitative and qualitative findings of the current pilot study support the scale-up of this work with minor modifications to the screening method as well as close monitoring and motivational interviewing to improve adherence in terms of physical activity.

TRIAL REGISTRATION NUMBER: NCT05928338.

PMID:37527890 | DOI:10.1136/bmjopen-2022-070913

BMJ Open. 2023 Aug 1;13(8):e074630. doi: 10.1136/bmjopen-2023-074630.

ABSTRACT

BACKGROUND: Epidemiological studies do not provide accurate statistics on the percentage of breastfeeding women experiencing sexual dysfunctions and restraining from sexual activity. The data vary between 40% and 83% in the first group and 20-50% in the second one. Despite excessive studies on contributors to intimacy changes, breast feeding received little attention from researchers. The relationship between lactation and postpartum sexual dysfunctions remains unclear. This systematic review and meta-analysis will synthesise available data and establish the link between breast feeding and sexuality problems.

METHODS AND ANALYSIS: A comprehensive literature search will be performed in biomedical databases PubMed/Medline, Scopus, Web of Science, EMBASE and CINAHL. We will extract peer-reviewed original studies written in English, Arabic or Polish from 2000 to June 2023. We will also search for reports from international health organisations and local health authorities. The preliminary search was performed on 04 April 2023. The studies must provide data on dysfunction prevalence/incidence and the strength of the relationship between breast feeding and sexuality in generally healthy women. The Covidence software will be used to perform literature screening, data extraction and quality assessment of individual studies. We will use a random-effects model meta-analysis to calculate pooled weighted frequency measures and effect size. Between-study heterogeneity will be assessed with the I² test.

ETHICS AND DISSEMINATION: This meta-analysis does not require ethical approval because it synthesises data from previously published original studies. The final work will be published in a peer-reviewed journal and presented at scientific conferences.

PROSPERO REGISTRATION NUMBER: CRD42023411053.

PMID:37527888 | DOI:10.1136/bmjopen-2023-074630

BMJ Open. 2023 Aug 1;13(8):e075736. doi: 10.1136/bmjopen-2023-075736.

ABSTRACT

INTRODUCTION: Friedreich ataxia (FA) is the most common hereditary ataxia in Europe, characterised by progressively worsening movement and speech impairments with a typical onset before the age of 25 years. The symptoms affect the patients’ health-related quality of life (HRQoL) and psychosocial health. FA leads to an increasing need for care, associated with an economic burden. Little is known about the impact of FA on daily lives and HRQoL. To fill that gap, we will assess patient-reported, psychosocial and economic outcomes using momentary data assessment via a mobile health application (app).

METHODS AND ANALYSIS: The PROFA Study is a prospective observational study. Patients with FA (n=200) will be recruited at six European study centres (Germany, France and Austria). We will interview patients at baseline in the study centre and subsequently assess the patients’ health at home via mobile health app. Patients will self-report ataxia severity, HRQoL, speech and hearing disabilities, coping strategies and well-being, health services usage, adverse health events and productivity losses due to informal care on a daily to monthly basis on the app for 6 months. Our study aims to (1) validate measurements of HRQoL and psychosocial health, (2) assess the usability of the mobile health app, and (3) use descriptive and multivariate statistics to analyse patient-reported and economic outcomes and the interaction effects between these outcomes. Insights into the app’s usability could be used for future studies using momentary data assessments to measure outcomes of patients with FA.

ETHICS AND DISSEMINATION: Ethical approval has been obtained from the Ethics Committee of the University Medicine of Greifswald, (BB096/22a, 26 October 2022) and from all local ethics committees of the participating study sites. Findings of the study will be published in peer-reviewed journals, presented at relevant international/national congresses and disseminated to German and French Patient Advocacy Organizations.

TRIAL REGISTRATION NUMBER: ClinicalTrials.gov Registry (NCT05943002); Pre-results.

PMID:37527887 | DOI:10.1136/bmjopen-2023-075736

Res Synth Methods. 2023 Aug 1. doi: 10.1002/jrsm.1660. Online ahead of print.

ABSTRACT

Missing data complicates statistical analyses in multi-site studies, especially when it is not feasible to centrally pool individual-level data across sites. We combined meta-analysis with within-site multiple imputation for one-step estimation of the average causal effect (ACE) of a target population comprised of all individuals from all data-contributing sites within a multi-site distributed data network, without the need for sharing individual-level data to handle missing data. We considered two orders of combination and three choices of weights for meta-analysis, resulting in six approaches. The first three approaches, denoted as RR + metaF, RR + metaR and RR + std, first combined results from imputed data sets within each site using Rubin’s rules and then meta-analyzed the combined results across sites using fixed-effect, random-effects and sample-standardization weights, respectively. The last three approaches, denoted as metaF + RR, metaR + RR and std + RR, first meta-analyzed results across sites separately for each imputation and then combined the meta-analysis results using Rubin’s rules. Simulation results confirmed very good performance of RR + std and std + RR under various missing completely at random and missing at random settings. A direct application of the inverse-variance weighted meta-analysis based on site-specific ACEs can lead to biased results for the targeted network-wide ACE in the presence of treatment effect heterogeneity by site, demonstrating the need to clearly specify the target population and estimand and properly account for potential site heterogeneity in meta-analyses seeking to draw causal interpretations. An illustration using a large administrative claims database is presented.

PMID:37527843 | DOI:10.1002/jrsm.1660

Stat Med. 2023 Aug 1. doi: 10.1002/sim.9862. Online ahead of print.

ABSTRACT

Natural history of hepatitis B or C is comprised of multiple milestones such as liver cirrhosis and liver cancer. To fully characterize its natural course, semicompeting risks represent a common problem where liver cirrhosis and liver cancer are both of interest, but only the former may be censored by the latter. Copula, frailty and multistate models serve as well-established analytics for semicompeting risks. Here, we cast the semicompeting risks in a mediation framework, with liver cirrhosis as a mediator and liver cancer as an outcome. We define the indirect and direct effects as the effects of an exposure on the liver cancer incidence mediated and not mediated through liver cirrhosis, respectively. With the estimands derived as conditional probabilities, we derive respective expressions under the copula, frailty, and multistate models. Next, we propose estimators based on nonparametric maximum likelihood or U-statistics and establish their asymptotic results. Numerical studies demonstrate that the efficiency of copula models leads to potential bias due to model misspecification. Moreover, the robustness of frailty models is accompanied by a loss in efficiency, and multistate models balance the efficiency and robustness. We demonstrate the utility of the proposed methods by a hepatitis study, showing that hepatitis B and C lead to a higher incidence of liver cancer by increasing liver cirrhosis incidence. Thus, mediation modeling provides a unified framework that accommodates various semicompeting risks models.

PMID:37527841 | DOI:10.1002/sim.9862

Am J Vet Res. 2023 Aug 5:1-7. doi: 10.2460/ajvr.23.05.0099. Online ahead of print.

ABSTRACT

OBJECTIVE: To evaluate antinociceptive efficacy of SC administration of hydromorphone hydrochloride and buprenorphine hydrochloride in ferrets (Mustela putorius furo).

ANIMALS: 14 healthy adult ferrets (6 neutered males, 8 spayed females).

METHODS: In a randomized, blind, controlled, complete crossover design, all 14 ferrets received a single, SC injection of hydromorphone low dose (0.1 mg/kg), hydromorphone high dose (0.2 mg/kg), buprenorphine low dose (0.02 mg/kg), buprenorphine high dose (0.04 mg/kg), or saline solution (0.2 mL/kg). Sedation and forelimb withdrawal latency from a noxious thermal stimulation were evaluated, and behavior was recorded for a total of 8 hours postinjection.

RESULTS: Compared to saline, administration of hydromorphone at 0.2 mg/kg resulted in an estimated increase of withdrawal latencies of 7.4 seconds (95% CI, 3.2 to 11.6) at 60 minutes, of 6.6 seconds (2.4 to 10.8) at 90 minutes, of 6.0 seconds (1.8 to 10.2) at 120 minutes, of 7.0 seconds (2.9 to 11.1) at 180 minutes, and of 4.5 seconds (0.5 to 8.6) at 240 minutes. These differences were statistically significant. Hydromorphone administered at a lower dose and buprenorphine at either dose did not increase withdrawal latencies compared to saline. Based on the sedation score used in this study, signs of sedation increased over time in a similar fashion with all treatments, including saline. Erratic dysphoric-like behaviors occurred in all groups except for saline.

CLINICAL RELEVANCE: SC administration of hydromorphone at a dose of 0.2 mg/kg provided antinociception from 1 to 4 hours postinjection. Further validation of sedation scores in ferrets is warranted.

PMID:37527831 | DOI:10.2460/ajvr.23.05.0099