Categories
Nevin Manimala Statistics

Review of Statistical Methods for Evaluating the Performance of Survival or Other Time-to-Event Prediction Models (from Conventional to Deep Learning Approaches)

Korean J Radiol. 2021 Jul 1. doi: 10.3348/kjr.2021.0223. Online ahead of print.

ABSTRACT

The recent introduction of various high-dimensional modeling methods, such as radiomics and deep learning, has created a much greater diversity in modeling approaches for survival prediction (or, more generally, time-to-event prediction). The newness of the recent modeling approaches and unfamiliarity with the model outputs may confuse some researchers and practitioners about the evaluation of the performance of such models. Methodological literacy to critically appraise the performance evaluation of the models and, ideally, the ability to conduct such an evaluation would be needed for those who want to develop models or apply them in practice. This article intends to provide intuitive, conceptual, and practical explanations of the statistical methods for evaluating the performance of survival prediction models with minimal usage of mathematical descriptions. It covers from conventional to deep learning methods, and emphasis has been placed on recent modeling approaches. This review article includes straightforward explanations of C indices (Harrell’s C index, etc.), time-dependent receiver operating characteristic curve analysis, calibration plot, other methods for evaluating the calibration performance, and Brier score.

PMID:34269532 | DOI:10.3348/kjr.2021.0223

Categories
Nevin Manimala Statistics

Prehospital whole blood reduces early mortality in patients with hemorrhagic shock

Transfusion. 2021 Jul;61 Suppl 1:S15-S21. doi: 10.1111/trf.16528.

ABSTRACT

BACKGROUND: Low titer O+ whole blood (LTOWB) is being increasingly used for resuscitation of hemorrhagic shock in military and civilian settings. The objective of this study was to identify the impact of prehospital LTOWB on survival for patients in shock receiving prehospital LTOWB transfusion.

STUDY DESIGN AND METHODS: A single institutional trauma registry was queried for patients undergoing prehospital transfusion between 2015 and 2019. Patients were stratified based on prehospital LTOWB transfusion (PHT) or no prehospital transfusion (NT). Outcomes measured included emergency department (ED), 6-h and hospital mortality, change in shock index (SI), and incidence of massive transfusion. Statistical analyses were performed.

RESULTS: A total of 538 patients met inclusion criteria. Patients undergoing PHT had worse shock physiology (median SI 1.25 vs. 0.95, p < .001) with greater reversal of shock upon arrival (-0.28 vs. -0.002, p < .001). In a propensity-matched group of 214 patients with prehospital shock, 58 patients underwent PHT and 156 did not. Demographics were similar between the groups. Mean improvement in SI between scene and ED was greatest for patients in the PHT group with a lower trauma bay mortality (0% vs. 7%, p = .04). No survival benefit for patients in prehospital cardiac arrest receiving LTOWB was found (p > .05).

DISCUSSION: This study demonstrated that trauma patients who received prehospital LTOWB transfusion had a greater improvement in SI and a reduction in early mortality. Patient with prehospital cardiac arrest did not have an improvement in survival. These findings support LTOWB use in the prehospital setting. Further multi-institutional prospective studies are needed.

PMID:34269467 | DOI:10.1111/trf.16528

Categories
Nevin Manimala Statistics

A distance-based living donor kidney education program for Black wait-list candidates: A feasibility randomized trial

Clin Transplant. 2021 Jul 16. doi: 10.1111/ctr.14426. Online ahead of print.

ABSTRACT

Addressing racial disparities in living donor kidney transplants (LDKT) among Black patients warrants innovative programs to improve living donation rates. The Living Organ Video Educated Donors (LOVED) program is a 2-arm, culturally-tailored, distance-based, randomized controlled feasibility trial. The group-based, 8-week program used peer-navigator led video chat sessions and web-app video education for Black kidney waitlisted patients from a United States southeastern state. Primary feasibility results for LOVED (n=24) and usual care (n=24) arms included LOVED program tolerability (i.e., 95.8% retention), program fidelity (i.e., 78.9% video education adherence and 72.1% video chat adherence). LDKT attitudinal and knowledge results favored the LOVED group where a statistically significant effect was reported over 6-months for willingness to approach strangers (estimate ± SE: -1.0 ± 0.55, F(1, 45.3) = 7.5, p = 0.009) and self-efficacy to advocate for a LDKT -0.81 ± 0.31, F(1, 45.9) = 15.2, p < 0.001. Estimates were improved but not statistically significant for willingness to approach family and friends, LDKT knowledge and concerns for living donors (all p’s >0.088). Secondary measures at 6 months showed an increase in calls for LOVED compared to usual care (p = 0.008) though no differences were found for transplant center evaluations or LDKTs. Findings imply that LOVED increased screening calls and attitudes to approach potential donors but feasibility outcomes found program materials require modification to increase adherence. NCT03599102 This article is protected by copyright. All rights reserved.

PMID:34269480 | DOI:10.1111/ctr.14426

Categories
Nevin Manimala Statistics

Cold-stored leukoreduced whole blood: Extending the time between donation and filtration has minimal impact on in vitro quality

Transfusion. 2021 Jul;61 Suppl 1:S131-S143. doi: 10.1111/trf.16540.

ABSTRACT

BACKGROUND: Leukoreduced whole blood (LR-WB) has received renewed attention as alternative to component-based transfusion in trauma. According to the manufacturer’s instructions, leukoreduction should be carried out within 8 h after collection. This study assessed impact of (1) WB collection bag, (2) LR filtration, and (3) timing of filtration on in vitro quality.

STUDY DESIGN AND METHODS: WB collected into different vendor bags was held at room temperature for <8 h or >16 h but <24 h prior to LR. In vitro quality was assessed before and after filtration, and throughout 3 weeks of storage at 4°C. Cell count and hemoglobin levels were determined by hematology analyzer, platelet activation, and responsiveness to ADP by surface expression of P-selectin by flow cytometry, hemolysis by HemoCue, and metabolic parameters by blood gas analyzer. Hemostatic properties were assessed by rotational thromboelastometry. Plasma protein activities and clotting times were determined by automated coagulation analyzer or quantitative immunoblotting.

RESULTS: Bag type had no impact on WB in vitro quality. LR by filtration had some impact, but is aligned with data in the literature. The time between donation and filtration resulted in some statistically significant differences in metabolic activity, platelet yield, platelet activation, and factor protein activity initially; however, these differences in in vitro quality attributes decreased throughout 21-day cold storage.

CONCLUSION: WB hold time showed only a minor impact on WB in vitro quality, so it may be possible for blood processing facilities to explore extended hold times prior to filtration in order to provide greater operational flexibility.

PMID:34269454 | DOI:10.1111/trf.16540

Categories
Nevin Manimala Statistics

Efficacy of the compensatory reserve measurement in an emergency department trauma population

Transfusion. 2021 Jul;61 Suppl 1:S174-S182. doi: 10.1111/trf.16498.

ABSTRACT

BACKGROUND: The Compensatory Reserve Measurement (CRM) is a novel method used to provide early assessment of shock based on arterial wave form morphology changes. We hypothesized that (1) CRM would be significantly lower in those trauma patients who received life-saving interventions compared with those not receiving interventions, and (2) CRM in patients who received interventions would recover after the intervention was performed.

STUDY DESIGN AND METHODS: We captured vital signs along with analog arterial waveform data from trauma patients meeting major activation criteria using a prospective study design. Study team members tracked interventions throughout their emergency department stay.

RESULTS: Ninety subjects met inclusion with 13 receiving a blood product and 10 a major airway intervention. Most trauma was blunt (69%) with motor vehicle collisions making up the largest proportion (37%) of injury mechanism. Patients receiving blood products had lower CRM values just prior to administration versus those who did not (50% versus 58%, p = .045), and lower systolic pressure (SBP; 95 versus 123 mmHg, p = .005), diastolic (DBP; 62 versus 79, p = .007), and mean arterial pressure (MAP; 75 versus 95, p = .006), and a higher pulse rate (HR; 101 versus 89 bpm, p = .039). Patients receiving an airway intervention had lower CRM values just prior to administration versus those who did not (48% versus 58%, p = .062); however, SBP, DBP, MAP, and HR were not statistically distinguishable (p ≥ .645).

CONCLUSIONS: Our results support our hypotheses that the CRM distinguished those patients who received blood or an airway intervention from those who did not, and increased appropriately after interventions were performed.

PMID:34269446 | DOI:10.1111/trf.16498

Categories
Nevin Manimala Statistics

Management of work stress in science education lecturers’ population using rational emotive occupational health coaching: Implication for educational evaluators

J Community Psychol. 2021 Jul 16. doi: 10.1002/jcop.22667. Online ahead of print.

ABSTRACT

INTRODUCTION: Nigerian university lecturers face a lot of works in the day-to-day discharge of their responsibilities as academics. This constitutes a lot of stress for them as documented in the literature. However, literature is scarce on how these lecturers manage their work stress. Thus, the researchers explored the efficacy of rational emotive occupational health coaching (REOHC) on the management of work stress among lecturers.

METHOD: A randomised controlled trial group experimental research design was adopted using a sample size of 84 lecturers in Science Education Departments of the sampled universities. An occupational stress index was used in the gathering of necessary data for the study. The REOHC treatment lasted for 12 weeks after which the participants were posttested and a follow-up measure followed after 3 months of the termination of the treatment. Repeated measures analysis of variance statistical approach was used to analyse the obtained data.

RESULTS: It was revealed that REOHC was significantly (p < 0.05) effective in the management of work stress in a population of science education lecturers in Nigeria.

CONCLUSION: REOHC enables science education lecturers to effectively manage their work stress to maximise their work outputs. This finding implicates the expertise of Educational Evaluators.

PMID:34269427 | DOI:10.1002/jcop.22667

Categories
Nevin Manimala Statistics

Quantification of anti-A of IgM or IgG isotype using three different methodologies

Transfusion. 2021 Jul;61 Suppl 1:S214-S222. doi: 10.1111/trf.16491.

ABSTRACT

BACKGROUND: Reliability of ABO-antibody measurement is important in the context of supplying low-titer ABO incompatible plasma-containing blood products. This study investigated the correlation of anti-A measurements between three different titer methodologies.

METHODS: Thirty-four blood group O individuals were included. IgM and IgG anti-A was measured by three different methods: (1) manual method (Bio-Rad ID-gel card), (2) automated method (Immucor NEO), (3) flow cytometry (FC) with calibration in molecules of equivalent fluorochrome (MEF). Data were log2 transformed to titer steps (TS) and log2 MEF, respectively. All three methods were benchmarked against the 14/300 WHO anti-A/anti-B standard reagent.

RESULTS: The correlation between the manual and automated methods was statistically significant for both IgM (Spearman’s rs = 0.89, p < .0001) and IgG (rs = 0.95, p < .0001). The mean TS difference between the manual and automated methods was 0.61 for IgM (p = .0033) and 2.1 for IgG (p < .0001). The manual method yielded IgM titer results that were generally 1 titer step higher than the automated method, whereas for the IgG titrations the difference was generally a median of 2 TS higher. The FC median log2 MEF level was significantly correlated with TS of IgG and IgM for both manual and automated agglutination-based titer methods (0.69 ≤ r2 ≤ 0.91). With the WHO standard reagent, the manual method produced the expected results while the automated method’s results were 1 TS lower for both IgM and IgG at all dilutions tested.

CONCLUSION: These results indicate that all three methods are suitable for measuring anti-A in group O whole blood.

PMID:34269431 | DOI:10.1111/trf.16491

Categories
Nevin Manimala Statistics

Safety profile of low-titer group O whole blood in pediatric patients with massive hemorrhage

Transfusion. 2021 Jul;61 Suppl 1:S8-S14. doi: 10.1111/trf.16456.

ABSTRACT

BACKGROUND: Low-titer Group O Whole Blood (LTOWB) is used with increasing frequency in adult and pediatric trauma and massive bleeding transfusion protocols. There is a risk of acute hemolytic reactions in non-group O recipients due to the passive transfusion of anti-A and anti-B in the LTOWB. This study investigated the hemolysis risk among pediatric recipients of LTOWB.

STUDY DESIGN AND METHODS: Blood bank records were queried for pediatric recipients of LTOWB between June 2016 and August 2020 and merged with clinical data. The primary outcome was laboratory evidence of hemolysis as manifested by changes in lactate dehydrogenase (LDH), haptoglobin, total bilirubin, reticulocyte count, potassium, and creatinine. Per protocol, these values were collected on hospital days 0-2 for recipients of LTOWB. Transfusion reactions were reported to the hospital’s blood bank.

RESULTS: Forty-seven children received LTOWB transfusion between 2016 and 2020; 21 were group O and 26 were non-group O. The groups were comparable in terms of the total volume of transfused blood products, demographics, and clinical outcomes. The most common indication for LTOWB transfusion was hemorrhagic shock due to trauma. There were no clinically or statistically significant differences in baseline, post-transfusion day 1, or post-transfusion day 2 hemolysis markers between the group O and non-group O LTOWB recipients. There were no adverse events or transfusion reactions reported.

DISCUSSION: Use of up to 40 ml/kg of LTOWB appears to be serologically safe for children in hemorrhagic shock.

PMID:34269441 | DOI:10.1111/trf.16456

Categories
Nevin Manimala Statistics

Statistical modeling of longitudinal medical cost trajectory: renal cell cancer care cost analyses

Biostatistics. 2020 Jul 16:kxab024. doi: 10.1093/biostatistics/kxab024. Online ahead of print.

ABSTRACT

Estimating the current cost of cancer care is important to health policy makers. An indispensable step in cost projection is to estimate cost trajectories from an incident cohort of cancer patients using longitudinal medical cost data, accounting for terminal events such as death, and right censoring due to loss of follow-up. Since the cost of cancer care and survival are correlated, a scientifically meaningful quantity for inference in this context is the mean cost trajectory conditional on survival. We propose a two-stage semiparametric approach to estimate the longitudinal cost trajectories from a joint model of longitudinal medical costs and survival. The longitudinal cost trajectories corresponding to various survival times form a bivariate surface in a triangular area. The cost trajectories are estimated using the tensor products of discretized measurement time and survival, as well as effective ridge penalties for data in 2D arrays. The proposed approach balances the practical considerations of model flexibility, statistical efficiency, and computational tractability. We used the proposed method to estimate the cost trajectories of renal cell cancer patients using the Surveillance, Epidemiology, and End Results-Medicare linked database.

PMID:34269395 | DOI:10.1093/biostatistics/kxab024

Categories
Nevin Manimala Statistics

The link between attitudes toward probationers and job burnout in Turkish probation officers

J Community Psychol. 2021 Jul 16. doi: 10.1002/jcop.22673. Online ahead of print.

ABSTRACT

The goal of the current study was to investigate individual-level factors associated with job burnout among probation officers (POs) and, specifically, to examine if attitudes toward probationers were linked with job burnout in the context of the recently established probation system in Turkey. Participants (N = 115) were recruited from a probation office in Istanbul. Job burnout was assessed via three components: emotional exhaustion, depersonalization, and professional accomplishment. Results of structural equation modeling indicated that more favorable attitudes toward probationers were related to a lower sense of depersonalization and higher experience of professional accomplishment. However, POs’ attitudes toward probationers were not associated with emotional exhaustion. Our findings are discussed in light of the present empirical literature on the contextual factors influential in job burnout. Practical implications for burnout prevention point to the potential effectiveness of working on attitudes among POs toward the people they supervise.

PMID:34269425 | DOI:10.1002/jcop.22673