Graph Co-Attentive Session-based Recommendation

Session-based recommendation aims to generate recommendations merely based on the ongoing session, which is a challenging task. Previous methods mainly focus on modeling the sequential signals or the transition relations between items in the current session using RNNs or GNNs to identify user’s intent for recommendation. Such models generally ignore the dynamic connections between the local and global item transition patterns, although the global information is taken into consideration by exploiting the global-level pair-wise item transitions. Moreover, existing methods that mainly adopt the cross-entropy loss with softmax generally face a serious over-fitting problem, harming the recommendation accuracy. Thus, in this article, we propose a Graph Co-Attentive Recommendation Machine (GCARM) for session-based recommendation. In detail, we first design a Graph Co-Attention Network (GCAT) to consider the dynamic correlations between the local and global neighbors of each node during the information propagation. Then, the item-level dynamic connections between the output of the local and global graphs are modeled to generate the final item representations. After that, we produce the prediction scores and design a Max Cross-Entropy (MCE) loss to prevent over-fitting. Extensive experiments are conducted on three benchmark datasets, i.e., Diginetica, Gowalla, and Yoochoose. The experimental results show that GCARM can achieve the state-of-the-art performance in terms of Recall and MRR, especially on boosting the ranking of the target item.

Safety, efficacy and delivery of isometric resistance training as an adjunct therapy for blood pressure control: a modified Delphi study

Uncontrolled hypertension remains the major risk factor for cardiovascular disease. Isometric resistance training (IRT) has been shown to be a useful nonpharmacological therapy for reducing blood pressure (BP); however, some exercise physiologists and other health professionals are uncertain of the efficacy and safety of IRT. Experts’ consensus was sought in light of the current variability of IRT use as an adjunct treatment for hypertension. An expert consensus-building analysis (Delphi study) was conducted on items relevant to the safety, efficacy and delivery of IRT. The study consisted of 3 phases: (1) identification of items and expert participants for inclusion; (2) a two-round modified Delphi exercise involving expert panelists to build consensus; and (3) a study team consensus meeting for a final item review. A list of 50 items was generated, and 42 international experts were invited to join the Delphi panel. Thirteen and 10 experts completed Delphi Rounds 1 and 2, respectively, reaching consensus on 26 items in Round 1 and 10 items in Round 2. The study team consensus meeting conducted a final item review and considered the remaining 14 items for the content list. A final list of 43 items regarding IRT reached expert consensus: 7/10 items on safety, 11/11 items on efficacy, 10/12 items on programming, 8/10 items on delivery, and 7/7 on the mechanism of action. This study highlights that while experts reached a consensus that IRT is efficacious as an antihypertensive therapy, some still have safety concerns, and there is also ongoing conjecture regarding optimal delivery.

Development of patient-reported outcomes item set to evaluate acute treatment toxicity to pelvic online magnetic resonance-guided radiotherapy

Background A new technology in cancer treatment, the MR-linac, provides online magnetic resonance-guided radiotherapy (MRgRT) that combines real-time visualization of the tumor and surrounding tissue with radiation therapy to deliver treatment more accurately. Online MRgRT makes it possible to minimize treatment volume, potentially reducing acute treatment toxicity. Patient-reported outcomes (PRO) add the patient perspective to evaluating treatment toxicity related to new technology. The objective of this mixed-methods study was to develop and explore the content validity of a set of PRO items to evaluate acute pelvic toxicity to radiotherapy including online MRgRT. Methods A literature review and chart audit were conducted to identify symptomatic adverse events (AEs) to be selected from the Patient-Reported Outcomes Version of the Common Terminology Criteria for Adverse Events (PRO-CTCAE) library and European Organisation for Research and Treatment of Cancer (EORTC) item library. To validate the content, the item set was applied in a prospective pilot cohort of patients referred for primary pelvic RT with curative intent. Patients reported symptoms weekly during RT (4–8 weeks) and the subsequent 4 weeks. Follow-up reports were collected at 8, 12, and 24 weeks after RT. To ensure symptom coverage clinician-reported toxicity and individual patient interviews were conducted. The symptomatic AEs were included in the final item set if ≥20% of patients reported them. Results Eighteen acute symptomatic AEs were selected for the initial item set. Forty patients (32 prostate cancer, 8 cervical cancer) were included in the pilot study. Patients with prostate cancer and those with cervical cancer both reported all 18 acute AEs. However, vomiting was not reported by > 20% of patients thus excluded from the item set. Adding a few diagnosis-specific AEs to the final item set was required for both prostate and cervical cancer patients. Conclusions A PRO item set for patients with pelvic cancer treated with radiotherapy with a curative intent was developed and content validity explored. In the pilot study, the item set captured the most common acute symptomatic AEs for patients with prostate and cervical cancer related to pelvic RT including online MRgRT. Further validation of the content in broader disease sites would be needed in future studies.

Overlapping as final-item completion in Mandarin conversation

This study locates as its focus the site for the final item in a sentence-in-progress as a late but systematic opportunity space for co-completing sentences by another speaker, and as a systematic site for brief overlaps. A second speaker may supply a version of the final item as projected by the grammatical structure of the sentence-so-far in given contexts to offer assistance for the searched-for final item upon the current speaker’s displayed delivery trouble, or to show an early recognition of what the current turn is doing and what it takes for its completion in the absence of any display of delivery trouble. The overlap in the first case may be ‘accidental’ when the first speaker is able to produce his/her own final item a moment later, or it may be an ‘achieved’ early start in the second case. The same opportunity space may also be ‘exploited’. Final items proposed by the second speaker may generate a local sequence where its acceptability becomes relevant. Post-overlap responses by the first speaker often show acceptance, sometimes with qualification. We argue that overlapping final-item completion is a result of speakers’ active participation and high involvement, and is motivated by the fundamental baseline of cooperation and collaboration in human social interaction.

Detecting unreported recalls in memory experiments

We investigate the role of discarded memory retrievals in experiments on free recall from lists of items. Such retrievals are not explicitly reported but testable predictions can be derived by assuming that they correlate with a delay in the timings of recall. We report on three situations where such a delay occurs: if the final item was already recalled (“silent recency effect”); if the item that, within the list, follows the latest recalled item was already recalled (“silent contiguity effect”); and in sequential recalls within highly performing trials (“sequential slowdown”). All these effects can be reproduced by a minimal model where the discarding of memories (“bouncing”) occurs either if they are repetitious or, in strategically organized trials, if they are not sequential. Based on our findings we propose various approaches to further probing the submerged dynamics of memory retrieval.

Development and validation of Malaysia Medication Adherence Assessment Tool (MyMAAT) for diabetic patients

Medication non-adherence remains a significant barrier in achieving better health outcomes for patients with chronic diseases. Previous self-reported medication adherence tools were not developed in the context of the Malaysia population. The most commonly used tool, MMAS-8, is no longer economical because it requires a license and currently every form used is charged. Hence, there is a need to develop and validate a new medication adherence tool. The Malaysia Medication Adherence Assessment Tool (MyMAAT) was developed by a multidisciplinary team with expertise in medication adherence and health literacy. The face and content validities of the MyMAAT was established by a panel of experts. A total of 495 patients with type 2 diabetes were recruited from the Ministry of Health facilities consisting of five hospitals and five primary health clinics. A test-retest was conducted on 42 of the patients one week following their first data collection. Exploratory factor analysis was performed to evaluate the validity of the MyMAAT. The final item for MyMAAT was compared with SEAMS, HbA1c%, Medication Possession ratio (MPR) score, and pharmacist’s subjective assessment for its hypothesis testing validity. The MyMAAT-12 achieved acceptable internal consistency (Cronbach’s alpha = 0.910) and stable reliability as the test-retest score showed good to excellent correlation (Spearman’s rho = 0.96, p = 0.001). The MyMAAT has significant moderate association with SEAMS (Spearman’s rho = 0.44, p = < 0.001) and significant relationship with HbA1c (< 8% and ≥ 8%) (χ2(1) = 13.4, p < 0.001), MPR (χ2(1) = 13.6, p < 0.001) and pharmacist’s subjective assessment categories (χ2(1) = 31, p < 0.001). The sensitivity of MyMAAT-12, tested against HbA1c% was 72.9% while its specificity was 43%. This study demonstrates that the MyMAAT-12 together with other methods of assessment may make a better screening tool to identify patients who were non-adherence to their medications.

Punctuation

The Punctuation chapter of the 11th edition of the AMA Manual of Style includes guidance on usage of many punctuation marks, as well as examples to illustrate the style. Included are the period, question mark, and exclamation point; comma, semicolon, and colon; hyphens and dashes; forward slash (virgule); parentheses and brackets; quotation marks; apostrophe; and ellipses. Expanded guidance on punctuation within lists or enumerations has been added, as have several policy changes: no hyphen in email and no period in the reference list after a URL or a DOI if either of these is the final item in the reference.

Development and Psychometric Evaluation of an Item Bank for Computerized Adaptive Testing of the EORTC Insomnia Dimension in Cancer Patients (EORTC CAT-SL)

To further advance assessment of patient-reported outcomes, the European Organisation of Research and Treatment of Cancer (EORTC) Quality of Life Group has developed computerized adaptive test (CAT) versions of all EORTC Quality of Life Core Questionnaire (QLQ-C30) scales/items. The aim of this study was to develop and evaluate an item bank for CAT measurement of insomnia (CAT-SL). In line with the EORTC guidelines, the developmental process comprised four phases: (I) defining the concept insomnia and literature search, (II) selection and formulation of new items, (III) pre-testing and (IV) field-testing, including psychometric analyses of the final item bank. In phase I, the literature search identified 155 items that were compatible with our conceptualisation of insomnia, including both quantity and quality of sleep. In phase II, following a multistep-approach, this number was reduced to 15 candidate items. Pre-testing of these items in cancer patients (phase III) resulted in an item list of 14 items, which were field-tested among 1094 patients in phase IV. Psychometric evaluations showed that eight items could be retained in a unidimensional model. The final item bank yielded greater measurement precision than the original QLQ-C30 insomnia item. It was estimated that administering two or more items from the insomnia item bank with CAT results in a saving in sample size between approximately 15–25%. The 8-item EORTC CAT-SL item bank facilitates precise and efficient measurement of insomnia as part of the EORTC CAT system of health-related quality life assessment in both clinical research and practice.

Anchoring Effects in Facial Attractiveness

First impressions from faces emerge quickly and shape subsequent behaviour. Given that different pictures of the same face evoke different impressions, we asked whether presentation order affects the overall impression of the person. In three experiments, we presented naturally varying photos of a person’s face in ascending (low-to-high) or descending (high-to-low) order of attractiveness. We found that attractiveness ratings for a subsequent test item were higher for the descending condition than for the ascending condition (Experiment 1), consistent with anchoring effects. In Experiment 2, we ruled out contrast between the final item and the test item as the cause of the effect by demonstrating anchoring within the sequence itself. In Experiment 3, we found that order of image presentation also affected dating decisions. Our findings demonstrate that first impressions from faces depend not only on visual information but also on the order in which that information is received. We suggest that models of impression formation and learning of individual faces could be improved by considering temporal order of encounters.

LO08: PROM-ED: the development and testing of a patient-reported outcome measure for use with emergency department patients who are discharged home

Introduction: Patient-reported outcome measures (PROM) are questionnaires that can be used to elicit care outcome information from patients. We sought to develop and validate the first PROM for adult patients without a primary mental health or addictions presentation receiving emergency department (ED) care and who were not hospitalized. Methods: PROM development used a multi-phase process based on national and international guidance (FDA, NQF, ISPOR). Phase 1: ED outcome conceptual framework qualitative interviews with ED patients post-discharge informed four core domains (previously published). Phase 2: Item generation scoping review of the literature and existing instruments identified candidate questions relevant for each domain for inclusion in tool. Phase 3: Cognitive debriefing existing and newly written questions were tested with ED patients post-discharge for comprehension and wording preference. Phase 4: Field and validity testing revised tool pilot tested on a national online survey panel and then again at 2 weeks (test-retest). Phase 5: Final item reduction using a Delphi process involving ED clinicians, researchers, patients and system administrators. Phase 6: Validation - psychometric testing of PROM-ED 1.0. Results: Four core outcome domains were defined in Phase 1: (1) understanding; (2) symptom relief; (3) reassurance and (4) having a plan. The domains informed a review of existing relevant questionnaires and instruments and the writing of additional questions creating an initial long-form questionnaire. Eight patients participated in cognitive debriefing of the long-form questionnaire. Expert clinicians, researchers and patient partners provided input on item refinement and reduction. Four hundred forty-four patients completed a second version of the long-form questionnaire (add in retest numbers) which informed the final item reduction process by a modified Delphi method involving 21 diverse contributors. The questionnaire was validated and underwent final revisions to create the 21 questions that constitute PROM-ED 1.0. Conclusion: Using accepted PROM instrument development methodology, we developed the first outcome questionnaire for use with adult ED patients who are not hospitalized. This questionnaire can be used to systematically gather patient-reported outcome information that could support and inform improvement work in ED care.