Nevin Manimala

Stat Methods Med Res. 2022 Nov 1:9622802221134172. doi: 10.1177/09622802221134172. Online ahead of print.

ABSTRACT

We propose a flexible and scalable approximate Bayesian inference methodology for the Cox Proportional Hazards model with partial likelihood. The model we consider includes nonlinear covariate effects and correlated survival times. The proposed method is based on nested approximations and adaptive quadrature, and the computational burden of working with the log-partial likelihood is mitigated through automatic differentiation and Laplace approximation. We provide two simulation studies to show the accuracy of the proposed approach, compared with the existing methods. We demonstrate the practical utility of our method and its computational advantages over Markov Chain Monte Carlo methods through the analysis of Kidney infection times, which are paired, and the analysis of Leukemia survival times with a semi-parametric covariate effect and spatial variation.

PMID:36317395 | DOI:10.1177/09622802221134172

Stat Methods Med Res. 2022 Nov 1:9622802221133554. doi: 10.1177/09622802221133554. Online ahead of print.

ABSTRACT

Survival data with multiple outcomes are frequently encountered in biomedical investigations. An illustrative example comes from Alzheimer’s Disease Neuroimaging Initiative study where the cognitively normal subjects may clinically progress to mild cognitive impairment and/or Alzheimer’s disease dementia. Transition time from normal cognition to mild cognitive impairment and that from mild cognitive impairment to Alzheimer’s disease are expected to be correlated within subjects and the dependence is often accommodated by the frailty (random effects). Estimation in the frailty model unavoidably involves multiple integrations which may be intractable and hence leads to severe computational challenges, especially in the presence of high-dimensional covariates. In this paper, we propose efficient minorization-maximization algorithms in the frailty model for survival data with multiple outcomes. The alternating direction method of multipliers is further incorporated for simultaneous variable selection and homogeneity pursuit via regularization and fusion. Extensive simulation studies are conducted to assess the performance of the proposed algorithms. An application to the Alzheimer’s Disease Neuroimaging Initiative data is also provided to illustrate their practical utilities.

PMID:36317365 | DOI:10.1177/09622802221133554

Stat Med. 2022 Nov 1. doi: 10.1002/sim.9599. Online ahead of print.

ABSTRACT

There is a growing interest in current medical research to develop personalized treatments using a molecular-based approach. The broad goal is to implement a more precise and targeted decision-making process, relative to traditional treatments based primarily on clinical diagnoses. Specifically, we consider patients affected by Acute Myeloid Leukemia (AML), an hematological cancer characterized by uncontrolled proliferation of hematopoietic stem cells in the bone marrow. Because AML responds poorly to chemotherapeutic treatments, the development of targeted therapies is essential to improve patients’ prospects. In particular, the dataset we analyze contains the levels of proteins involved in cell cycle regulation and linked to the progression of the disease. We evaluate treatment effects within a causal framework represented by a Directed Acyclic Graph (DAG) model, whose vertices are the protein levels in the network. A major obstacle in implementing the above program is represented by individual heterogeneity. We address this issue through a Dirichlet Process (DP) mixture of Gaussian DAG-models where both the graphical structure as well as the allied model parameters are regarded as uncertain. Our procedure determines a clustering structure of the units reflecting the underlying heterogeneity, and produces subject-specific estimates of causal effects based on Bayesian Model Averaging (BMA). With reference to the AML dataset, we identify different effects of protein regulation among individuals; moreover, our method clusters patients into groups that exhibit only mild similarities with traditional categories based on morphological features.

PMID:36317356 | DOI:10.1002/sim.9599

Bone Joint J. 2022 Nov;104-B(11):1196-1201. doi: 10.1302/0301-620X.104B11.BJJ-2022-0194.R2.

ABSTRACT

AIMS: Although CT is considered the benchmark to measure femoral version, 3D biplanar radiography (hipEOS) has recently emerged as a possible alternative with reduced exposure to ionizing radiation and shorter examination time. The aim of our study was to evaluate femoral stem version in postoperative total hip arthroplasty (THA) patients and compare the accuracy of hipEOS to CT. We hypothesize that there will be no significant difference in calculated femoral stem version measurements between the two imaging methods.

METHODS: In this study, 45 patients who underwent THA between February 2016 and February 2020 and had both a postoperative CT and EOS scan were included for evaluation. A fellowship-trained musculoskeletal radiologist and radiological technician measured femoral version for CT and 3D EOS, respectively. Comparison of values for each imaging modality were assessed for statistical significance.

RESULTS: Comparison of the mean postoperative femoral stem version measurements between CT and 3D hipEOS showed no significant difference (p = 0.862). In addition, the two version measurements were strongly correlated (r = 0.95; p < 0.001), and the mean paired difference in postoperative femoral version for CT scan and 3D biplanar radiography was -0.09° (95% confidence interval -1.09 to 0.91). Only three stem measurements (6.7%) were considered outliers with a > 5° difference.

CONCLUSION: Our study supports the use of low-dose biplanar radiography for the postoperative assessment of femoral stem version after THA, demonstrating high correlation with CT. We found no significant difference for postoperative femoral version when comparing CT to 3D EOS. We believe 3D EOS is a reliable option to measure postoperative femoral version given its advantages of lower radiation dosage and shorter examination time.Cite this article: Bone Joint J 2022;104-B(11):1196-1201.

PMID:36317354 | DOI:10.1302/0301-620X.104B11.BJJ-2022-0194.R2

Bone Joint J. 2022 Nov;104-B(11):1256-1265. doi: 10.1302/0301-620X.104B11.BJJ-2022-0653.R1.

ABSTRACT

AIMS: To determine whether platelet-rich plasma (PRP) injection improves outcomes two years after acute Achilles tendon rupture.

METHODS: A randomized multicentre two-arm parallel-group, participant- and assessor-blinded superiority trial was undertaken. Recruitment commenced on 28 July 2015 and two-year follow-up was completed in 21 October 2019. Participants were 230 adults aged 18 years and over, with acute Achilles tendon rupture managed with non-surgical treatment from 19 UK hospitals. Exclusions were insertion or musculotendinous junction injuries, major leg injury or deformity, diabetes, platelet or haematological disorder, medication with systemic corticosteroids, anticoagulation therapy treatment, and other contraindicating conditions. Participants were randomized via a central online system 1:1 to PRP or placebo injection. The main outcome measure was Achilles Tendon Rupture Score (ATRS) at two years via postal questionnaire. Other outcomes were pain, recovery goal attainment, and quality of life. Analysis was by intention-to-treat.

RESULTS: A total of 230 participants were randomized, 114 to PRP and 116 to placebo. Two-year questionnaires were sent to 216 participants who completed a six-month questionnaire. Overall, 182/216 participants (84%) completed the two-year questionnaire. Participants were aged a mean of 46 years (SD 13.0) and 25% were female (57/230). The majority of participants received the allocated intervention (219/229, 96%). Mean ATRS scores at two years were 82.2 (SD 18.3) in the PRP group (n = 85) and 83.8 (SD 16.0) in the placebo group (n = 92). There was no evidence of a difference in the ATRS at two years (adjusted mean difference -0.752, 95% confidence interval -5.523 to 4.020; p = 0.757) or in other secondary outcomes, and there were no re-ruptures between 24 weeks and two years.

CONCLUSION: PRP injection did not improve patient-reported function or quality of life two years after acute Achilles tendon rupture compared with placebo. The evidence from this study indicates that PRP offers no patient benefit in the longer term for patients with acute Achilles tendon rupture.Cite this article: Bone Joint J 2022;104-B(11):1256-1265.

PMID:36317349 | DOI:10.1302/0301-620X.104B11.BJJ-2022-0653.R1

Bone Joint J. 2022 Nov;104-B(11):1234-1241. doi: 10.1302/0301-620X.104B11.BJJ-2022-0422.R2.

ABSTRACT

AIMS: This study compared patients who underwent arthroscopic repair of large to massive rotator cuff tears (LMRCTs) with isolated incomplete repair of the tear and patients with incomplete repair with biceps tendon augmentation. We aimed to evaluate the additional benefit on clinical outcomes and the capacity to lower the re-tear rate.

METHODS: We retrospectively reviewed 1,115 patients who underwent arthroscopic rotator cuff repair for full-thickness tears between October 2011 and May 2019. From this series, we identified 77 patients (28 male, 49 female) with a mean age of 64.1 years (50 to 80). Patients were classified into groups A (n = 47 incomplete) and B (n = 30 with biceps augmentation) according to the nature of their reconstruction. Clinical scores were checked preoperatively and at six months, one year, and two years postoperatively. In preoperative MRI, we measured the tear size, the degree of fatty infiltration, and muscle volume ratio of the supraspinatus. In postoperative MRI, the integrity of the repaired rotator cuff tendon was assessed using the Sugaya classification. Tendon thickness at the footprint was evaluated on T2-weighted oblique coronal view.

RESULTS: There were no significant differences in the initial preoperative demographic characteristics. In both groups, there were significant improvements in postoperative clinical scores (p < 0.001). However, most clinical outcomes, including range of motion measurements (forward elevation, external rotation, internal rotation, and abduction), showed no differences between the pre- and postoperative values. Comparing the postoperative outcomes of both groups, no further improvement from biceps augmentation was found. Group B, although not reaching statistical significance, had more re-tears than group A (30% vs 15%; p = 0.117).

CONCLUSION: In LMRCTs, biceps augmentation provided no significant improvement of an incomplete repair. Therefore, biceps augmentation is not recommended in the treatment of LMRCTs.Cite this article: Bone Joint J 2022;104-B(11):1234-1241.

PMID:36317346 | DOI:10.1302/0301-620X.104B11.BJJ-2022-0422.R2

Bone Joint J. 2022 Nov;104-B(11):1225-1233. doi: 10.1302/0301-620X.104B11.BJJ-2022-0386.R1.

ABSTRACT

AIMS: The aim of this study was to compare the cost-effectiveness of surgical fixation with Kirschner (K-)wire ersus moulded casting after manipulation of a fracture of the distal radius in an operating theatre setting.

METHODS: An economic evaluation was conducted based on data collected from the Distal Radius Acute Fracture Fixation Trial 2 (DRAFFT2) multicentre randomized controlled trial in the UK. Resource use was collected at three, six, and 12 months post-randomization using trial case report forms and participant-completed questionnaires. Cost-effectiveness was reported in terms of incremental cost per quality-adjusted life year (QALY) gained from an NHS and personal social services perspective. Sensitivity analyses were conducted to examine the robustness of cost-effectiveness estimates, and decision uncertainty was handled using confidence ellipses and cost-effectiveness acceptability curves.

RESULTS: In the base case analysis, surgical fixation with K-wire was more expensive (£29.65 (95% confidence interval (CI) -94.85 to 154.15)) and generated lower QALYs (0.007 (95% CI -0.03 to 0.016)) than moulded casting, but this difference was not statistically significant. The probability of K-wire being cost-effective at a £20,000 per QALY cost-effectiveness threshold was 24%. The cost-effectiveness results remained robust in the sensitivity analyses.

CONCLUSION: The findings suggest that surgical fixation with K-wire is unlikely to be a cost-effective alternative to a moulded cast in adults, following manipulation of a fracture of the distal radius in a theatre setting.Cite this article: Bone Joint J 2022;104-B(11):1225-1233.

PMID:36317342 | DOI:10.1302/0301-620X.104B11.BJJ-2022-0386.R1

Int J Geriatr Psychiatry. 2022 Dec;37(12). doi: 10.1002/gps.5820.

ABSTRACT

OBJECTIVES: The aim was to confirm the relationship between social support (SS), post-stroke depression (PSD), and quality of life (QOL) and determine the mediating effect of PSD in stroke survivors. Additionally, we tested the impact of economic development level (in developed or developing countries) on the relationship between SS, PSD, and QOL.

METHODS: This study utilized meta-analytic structural equation modeling on systematically searched articles from PubMed, MEDLINE, Cochrane Library, Scopus, PsycINFO, Web of Science, China National Knowledge Infrastructure, and WanFang data published from inception to February 2022. Collect the sample size (n) of each study and the associations of observed variables, and conduct meta-analysis path analysis using AMOS 23.0 to assess the relationships. Concurrently, the effects of the national economic development level were extracted for moderator analysis.

RESULTS: A total of 28 studies (N = 3967) were included for analysis. SS and PSD were significant predictors of QOL (both p < 0.01). PSD mediated the relationship between SS and QOL (β = 0.31; 95% confidence interval 0.273-0.345; p < 0.01). Furthermore, in developed countries, SS was not statistically correlated with PSD (p = 0.811) compared to developing countries.

CONCLUSIONS: Improving SS may help improve the QOL of stroke patients. PSD should be addressed using a comprehensive approach that includes interventions to enhance the QOL. Additionally, people have different psychological reactions to SS at different stages of economic development; thus, further research is needed to develop different measurement standards for patients according to the country’s level of economic development.

PMID:36317324 | DOI:10.1002/gps.5820

Physiol Meas. 2022 Oct 28. doi: 10.1088/1361-6579/ac9e8a. Online ahead of print.

ABSTRACT

BACKGROUND: Clinical medicine relies heavily on the synthesis of information and data from multiple sources. However, often simple feature concatenation is used as a strategy for developing a multimodal machine learning model in the cardiovascular domain, and thus the models are often limited by pre-selected features and moderate accuracy.

METHOD: We proposed a two-branched joint fusion model for fusing the 12-lead electrocardiogram (ECG) signal data with clinical variables from the electronic medical record (EMR) in an end-to-end deep learning architecture. The model follows the joint fusion scheme and learns complementary information from ECG and EMR. Retrospective data from the Mayo Clinic Health Systems across four sites (La Crosse, WI; Mankato, MN; Rochester, MN; Scottsdale, AZ) for patients that underwent percutaneous coronary intervention (PCI) between January 2006 and December 2018 were obtained. Model performance was assessed by area under the receiver-operating characteristics (AUROC) and Delong’s test.

RESULTS: The final cohort included 17,356 unique patients with a mean age of 67.2 ± 12.6 and 9,163 (52.7%) were male. The joint fusion model outperformed the ECG time-domain model with statistical margin. The model with clinical data obtained the highest AUROC for all-cause mortality (0.91 at 6 months) but the joint fusion model outperformed for cardiovascular outcomes – heart failure hospitalization and ischemic stroke with a significant margin (Delong’s p < 0.05).

CONCLUSION: To the best of our knowledge, this is the first study that developed a deep learning model with joint fusion architecture for the prediction of post-PCI prognosis and outperformed machine learning models developed using traditional single-source features (clinical variables or ECG features). Adding ECG data with clinical variables did not improve prediction of all-cause mortality as may be expected, but the improved performance of related cardiac outcomes shows that the fusion of ECG generates additional value.

PMID:36317320 | DOI:10.1088/1361-6579/ac9e8a

J Small Anim Pract. 2022 Oct 31. doi: 10.1111/jsap.13573. Online ahead of print.

ABSTRACT

OBJECTIVES: To assess treatment satisfaction and owner preference for two delivery devices (VetPen, MSD Animal Health, and U40 insulin syringes) and the effect on glycaemic control in diabetic dogs treated with porcine insulin zinc suspension.

MATERIALS AND METHODS: Randomised prospective cross-over study with two arms, each of 8 weeks. Twenty client-owned diabetic dogs on insulin treatment by U40 syringe were enrolled. Dogs were randomly assigned to receive insulin by syringe or pen injector for 2 months, followed by 2 months of the other injection method. Treatment satisfaction and owners’ insulin delivery device preference were assessed using a questionnaire. Glycaemic control was assessed using a clinical score, serum fructosamine and glycated haemoglobin (HbA1c%) at the time of the enrolment (T0) and the end of each arm of treatment (T2 and T4).

RESULTS: Treatment satisfaction differed for the two types of the delivery device when the order that each device was used was taken into consideration. Owners who used the syringe first did not have a significant preference for an injection device. In contrast, owners who used the pen injector first expressed a significant preference for VetPen compared to syringes. No significant differences in the number of dogs of Groups 1 and 2 with good and poor glycaemic control at T2 and T4 were detected.

CLINICAL SIGNIFICANCE: Overall treatment satisfaction and preference for the two delivery methods were similar. However, VetPen was preferred by owners who were randomised to use this device first. Glycaemic control did not appear to be affected by the insulin delivery device used.

PMID:36316285 | DOI:10.1111/jsap.13573