A binary classification problem is common in medical field, and we often use sensitivity, specificity, accuracy, negative and positive predictive values as measures of performance of a binary predictor. In computer science, a classifier is usually evaluated with precision (positive predictive value) and recall (sensitivity). As a single summary measure of a classifier’s performance, F₁ score, defined as the harmonic mean of precision and recall, is widely used in the context of information retrieval and information extraction evaluation since it possesses favorable characteristics, especially when the prevalence is low. Some statistical methods for inference have been developed for the F₁ score in binary classification problems; however, they have not been extended to the problem of multi-class classification. There are three types of F₁ scores, and statistical properties of these F₁ scores have hardly ever been discussed. We propose methods based on the large sample multivariate central limit theorem for estimating F₁ scores with confidence intervals.

Journal of Medical Internet Research - International Scientific Journal for Medical Research, Information and Communication on the Internet #Preprint #PeerReviewMe: Warning: This is a unreviewed preprint. Readers are warned that the document has not been peer-reviewed by expert/patient reviewers or an academic editor, may contain misleading claims, and is likely to undergo changes before final publication, if accepted, or may have been rejected/withdrawn. Readers with interest and expertise are encouraged to sign up as peer-reviewer, if the paper is within an open peer-review period. Please cite this preprint only for review purposes or for grant applications and CVs (if you are the author). Background: Human Papillomavirus (HPV) vaccine hesitancy among parents contributes to low vaccination coverage in adolescents. In order to improve healthcare provider communication and vaccine recommendation practices with hesitant parents, it is important to understand how providers perceive parental HPV vaccine hesitancy. Objective: This study aimed to characterize perceived reasons for parental HPV vaccine hesitancy and identify factors associated with perceived parental hesitancy among providers at community-based pediatric clinics. Methods: In 2018, providers in 23 community-based pediatric clinics in Tennessee were invited to complete an online baseline survey as part of a larger quality improvement study focused on HPV vaccine uptake. These survey data were used for a cross-sectional, secondary data analysis. Scale scores ranging from 0 to 100 were calculated for provider self-efficacy (confidence in ability to recommend HPV vaccine), provider outcome expectations (expectations that recommendation will influence parents’ decisions), and perceived parental HPV vaccine hesitancy. Provider confidence in HPV vaccine safety and effectiveness were categorized as high versus low. Clinic-level exposures examined were clinic size and rural/urban location. Descriptive analyses were used to characterize perceived parental barriers by provider type. Mixed-effects linear regression models were fit taking one exposure variable at a time, while controlling for provider type, age, gender, and race to identify provider- and clinic level factors associated with perceived parental barriers to HPV vaccination. Results: Of the 187 providers located in the 23 clinics, 137 completed the survey. The majority of physician providers were white and female, with a higher percentage of females among nurse practitioners and physician assistants. The most common parental barriers to HPV vaccination perceived by providers were concerns about HPV vaccine safety (88%), child being too young (78%), low risk of HPV infection for child through sexual activity (70%), and mistrust in vaccines (59%). In adjusted mixed models, perceived parental HPV vaccine hesitancy was significantly associated with several provider-level factors: Self-efficacy (P = 0.001), Outcome expectations (P < 0.001), and Confidence in HPV vaccine safety (P = 0.009). No significant associations were observed between perceived parental HPV vaccine hesitancy and clinic level factors Clinic size nor Location. Conclusions: Researchers developing provider-focused interventions to reduce parental HPV vaccine hesitancy should consider addressing providers’ self-efficacy, outcome expectations, and confidence in HPV vaccine safety to help providers communicate more effectively with HPV vaccine hesitant parents.

To evaluate the relative efficacy of novel retinoblastoma treatments, eye classification-specific success rates for current standard-of-care intravenous chemotherapy regimens must be known. This meta-analysis included studies if: (1) patients received intravenous chemotherapy for retinoblastoma, (2) globe salvage data was reported, (3) only intravenous chemoreduction (with/without local consolidation) was used. The outcome measure was globe salvage success without need for salvage radiotherapy, subdivided by disease classification and chemotherapy regimen. Data from 27 studies (1483 eyes) were pooled. By Reese–Ellsworth classification, globe salvage rates were 85% (95%CI:73–92%) for Group I, 78% (95%CI:70–85%) for Group II, 68% (95%CI:56–78%) for Group III, 47% (95%CI:34–60%) for Group IV, and 35% (95%CI:26–45%) for Group V (Va: 35% [95%CI:21–54%]; Vb: 42% [95%CI:29–56%]; those without sub-classification: 31% [95%CI:19–47%]). By International Classification, globe salvage rates were 93% (95%CI:80–97%) for Group A, 83% (95%CI:73–89%) for Group B, 73% (95%CI:54–86%) for Group C, 40% (95%CI:31–51%) for Group D, and 19% (95%CI:5–50%) for Group E. Standard carboplatin-etoposide-vincristine out-performed two-drug regimens (odds ratio (OR) = 1.9 (95%CI:1.3–3.0) for Groups I-IV and OR = 2.1 (95%CI:1.3–3.4) for Group V; p = 0.002 for each). For eyes with diffuse vitreous seeds (Vb), an enhanced regimen out-performed standard chemotherapy (OR = 2.4 [95%CI:1.3–4.7]; p = 0.004). In conclusion, two-drug regimens were less effective for all eyes, whereas enhanced regimens were more effective for eyes with vitreous seeds. Novel therapies can now be compared to these baseline globe salvage rates.

Objective: Changing Behavior through Physical Therapy (CBPT), a cognitive-behavioral–based program, has been shown to improve outcomes after lumbar spine surgery in patients with a high psychosocial risk profile; however, little is known about potential mechanisms associated with CBPT treatment effects. The purpose of this study was to explore potential mediators underlying CBPT efficacy after spine surgery.Methods: In this secondary analysis, 86 participants were enrolled in a randomized trial comparing a postoperative CBPT (n = 43) and education program (n = 43). Participants completed validated questionnaires at 6 weeks (baseline) and 3 and 6 months following surgery for back pain (Brief Pain Inventory), disability (Oswestry Disability Index), physical health (12-Item Short-Form Health Survey), fear of movement (Tampa Scale for Kinesiophobia), pain catastrophizing (Pain Catastrophizing Scale), and pain self-efficacy (Pain Self-Efficacy Questionnaire). Parallel multiple mediation analyses using Statistical Package for the Social Sciences (SPSS) were conducted to examine whether 3- and 6-month changes in fear of movement, pain catastrophizing, and pain self-efficacy mediate treatment outcome effects at 6 months.Results: Six-month changes, but not 3-month changes, in fear of movement and pain self-efficacy mediated postoperative outcomes at 6 months. Specifically, changes in fear of movement mediated the effects of CBPT treatment on disability (indirect effect = −2.0 [95% CI = −4.3 to 0.3]), whereas changes in pain self-efficacy mediated the effects of CBPT treatment on physical health (indirect effect = 3.5 [95% CI = 1.2 to 6.1]).Conclusions: This study advances evidence on potential mechanisms underlying cognitive-behavioral strategies. Future work with larger samples is needed to establish whether these factors are a definitive causal mechanism.Impact: Fear of movement and pain self-efficacy may be important mechanisms to consider when developing and testing psychologically informed physical therapy programs.

Adipocyte iron overload is a maladaptation associated with obesity and insulin resistance. The objective of the current study was to determine whether and how adipose tissue macrophages (ATMs) regulate adipocyte iron concentrations and whether this is impacted by obesity. Using bone marrow-derived macrophages (BMDMs) polarized to M0, M1, M2, or metabolically activated (MMe) phenotypes, we showed that MMe BMDMs and ATMs from obese mice have reduced expression of several iron-related proteins. Furthermore, the bioenergetic response to iron in obese ATMs was hampered. ATMs from iron-injected lean mice increased their glycolytic and respiratory capacities, thus maintaining metabolic flexibility, while ATMs from obese mice did not. Using an isotope-based system, we found that iron exchange between BMDMs and adipocytes was regulated by macrophage phenotype. At the end of the co-culture, MMe macrophages transferred and received more iron from adipocytes than M0, M1, and M2 macrophages. This culminated in a decrease in total iron in MMe macrophages and an increase in total iron in adipocytes compared with M2 macrophages. Taken together, in the MMe condition, the redistribution of iron is biased toward macrophage iron deficiency and simultaneous adipocyte iron overload. These data suggest that obesity changes the communication of iron between adipocytes and macrophages and that rectifying this iron communication channel may be a novel therapeutic target to alleviate insulin resistance.