Machine Learning Predicts the Fall Risk of Total Hip Arthroplasty Patients Based on Wearable Sensor Instrumented Performance Tests

BackgroundThe prevalence of falls affects the wellbeing of aging adults and places an economic burden on the healthcare system. Integration of wearable sensors into existing fall risk assessment tools enables objective data collection that describes the functional ability of patients. In this study, supervised machine learning was applied to sensor-derived metrics to predict the fall risk of patients following total hip arthroplasty.MethodsAt preoperative, 2-week, and 6-week postoperative appointments, patients (n = 72) were instrumented with sensors while they performed the timed-up-and-go walking test. Preoperative and 2-week postoperative data were used to form the feature sets and 6-week total times were used as labels. Support vector machine and linear discriminant analysis classifier models were developed and tested on various combinations of feature sets and feature reduction schemes. Using a 10-fold leave-some-subjects-out testing scheme, the accuracy, sensitivity, specificity, and area under the receiver-operator curve (AUC) were evaluated for all models.ResultsA high performance model (accuracy = 0.87, sensitivity = 0.97, specificity = 0.46, AUC = 0.82) was obtained with a support vector machine classifier using sensor-derived metrics from only the preoperative appointment. An overall improved performance (accuracy = 0.90, sensitivity = 0.93, specificity = 0.59, AUC = 0.88) was achieved with a linear discriminant analysis classifier when 2-week postoperative data were added to the preoperative data.ConclusionThe high accuracy of the fall risk prediction models is valuable for patients, clinicians, and the healthcare system. High-risk patients can implement preventative measures and low-risk patients can be directed to enhanced recovery care programs.     

Multi-Disciplinary Trauma Evaluation and Management Simulation (MD-TEAMS) training for emergency medicine and general surgery residents

BackgroundSuccessful trauma resuscitation relies on multi-disciplinary collaboration. In most academic programs, general surgery (GS) and emergency medicine (EM) residents rarely train together before functioning as a team.MethodsIn our Multi-Disciplinary Trauma Evaluation and Management Simulation (MD-TEAMS), EM and GS residents completed manikin-based trauma scenarios and were evaluated on resuscitation and communication skills. Residents were surveyed on confidence surrounding training objectives.ResultsResidents showed improved confidence running trauma scenarios in multi-disciplinary teams. Residents received lower communication scores from same-discipline vs cross-discipline faculty. EM residents scored higher in evaluation and planning domains; GS residents scored higher in action processes; groups scored equally in team management. Strong correlation existed between team leader communication and resuscitative skill completion.ConclusionMD-TEAMS demonstrated correlation between communication and resuscitation checklist item completion and communication differences by resident specialty. In the future, we plan to evaluate training-related resident behavior changes and specialty-specific communication differences by residents.     

The First Genome-Wide Association Study for Type 2 Diabetes in Youth: The Progress in Diabetes Genetics in Youth (ProDiGY) Consortium

The prevalence of type 2 diabetes in youth has increased substantially, yet the genetic underpinnings remain largely unexplored. To identify genetic variants predisposing to youth-onset type 2 diabetes, we formed ProDiGY, a multiethnic collaboration of three studies (TODAY, SEARCH, and T2D-GENES) with 3,006 youth case subjects with type 2 diabetes (mean age 15.1 ± 2.9 years) and 6,061 diabetes-free adult control subjects (mean age 54.2 ± 12.4 years). After stratifying by principal component–clustered ethnicity, we performed association analyses on ∼10 million imputed variants using a generalized linear mixed model incorporating a genetic relationship matrix to account for population structure and adjusting for sex. We identified seven genome-wide significant loci, including the novel locus rs10992863 in PHF2 (P = 3.2 × 10⁻⁸; odds ratio [OR] = 1.23). Known loci identified in our analysis include rs7903146 in TCF7L2 (P = 8.0 × 10⁻²⁰; OR 1.58), rs72982988 near MC4R (P = 4.4 × 10⁻¹⁴; OR 1.53), rs200893788 in CDC123 (P = 1.1 × 10⁻¹²; OR 1.32), rs2237892 in KCNQ1 (P = 4.8 × 10⁻¹¹; OR 1.59), rs937589119 in IGF2BP2 (P = 3.1 × 10⁻⁹; OR 1.34), and rs113748381 in SLC16A11 (P = 4.1 × 10⁻⁸; OR 1.04). Secondary analysis with 856 diabetes-free youth control subjects uncovered an additional locus in CPEB2 (P = 3.2 × 10⁻⁸; OR 2.1) and consistent direction of effect for diabetes risk. In conclusion, we identified both known and novel loci in the first genome-wide association study of youth-onset type 2 diabetes.     

Improving care through standardized treatment of spontaneous pneumothorax

PurposeThe objective of this quality improvement (QI) initiative was to implement a standardized clinical treatment protocol for patients presenting with primary spontaneous pneumothorax (PSP) in order to decrease hospital length of stay (LOS), diagnostic radiation exposure, and related cost.MethodsBaseline data from patients admitted with PSP from January 1, 2016 to July 31, 2018 were compared to data from patients managed using a newly developed evidence-based treatment pathway from August 1, 2018 to December 31, 2019. Standard QI methodology was used to track results.ResultsFifty-six episodes of PSP were observed during the baseline period and 40 episodes of PSP following initiation of the PSP protocol. The average LOS decreased from 4.5 days to 2.9 days. Patients underwent an average of 8.8 X-rays per admission preintervention versus 5.9 postintervention. The rate of CT scans decreased from 45% to 15% (p = 0.002). There was no significant difference in the rates of 30-day recurrence between the preintervention (13%) and postintervention (10%) groups (p = 0.7). Average admission costs per patient decreased by $1322 after adoption of the pathway.ConclusionsAdoption of a standardized treatment protocol for PSP led to a reduction in LOS, diagnostic imaging utilization, and cost without increasing clinical recurrence.Type of studyQuality improvement.Level of evidenceLevel III.     

Association of hospital and surgeon operative volumes and surgeon pediatric subspecialization with pediatric laparoscopic cholecystectomy outcomes: A population-based cohort study

BackgroundMost pediatric cholecystectomies are performed by adult general surgeons, but it is unclear whether outcomes differ by surgeon pediatric subspecialization, hospital procedure volume, or surgeon procedure volume. We aimed to determine whether higher hospital or surgeon laparoscopic cholecystectomy (LC) volume or surgeon pediatric subspecialization is associated with better outcomes after LC in children.MethodsWe performed a retrospective cohort study using statewide hospital discharge databases from the states of Florida, Georgia, and Iowa. We included children aged 4–18 years who underwent inpatient or outpatient LC for acute or chronic gallbladder disease in January 2010–August 2015.Propensity score weighting was used to estimate relationships between operative volumes or surgeon pediatric subspecialization and rates of readmission or emergency department (ED) visit within 30 days.ResultsA total of 5391 children were included (mean age 15.9 years, 81.6% female). Children operated on by surgeons with high LC volumes in hospitals with high LC volumes were less likely to experience a readmission or ED visit within 30 days (10.8% vs. 13.7%, p = 0.04). Additionally, children operated on by adult general surgeons in hospitals with high LC volumes were less likely to experience a readmission or ED visit within 30 days (10.9% vs. 13.8%, p = 0.03).ConclusionsChildren are less likely to be readmitted to the hospital or present to the ED after laparoscopic cholecystectomy if they receive their care from adult general surgeons at hospitals that frequently perform this procedure in both adults and children.Level of evidenceIII     

Management of cervicofacial lymphatic malformations requires a multidisciplinary approach

Background/PurposeCervicofacial lymphatic malformations (CFLM) are rare, potentially life-threatening vascular anomalies, yet reports on multidisciplinary treatment strategies are lacking. We evaluated outcomes for CFLMs following sclerotherapy, surgical resection, and/or medical management.MethodsWe identified children with a CFLM at a vascular anomalies center from 2004 to 2019. Exclusion criteria: retro-orbital malformations, untreated malformations, patients without follow-up. Primary clinical outcome was contour improvement, with significance defined as LM volume reduction of > 50% by cross-sectional imaging.ResultsSixty-three children met inclusion criteria: 35 with macrocystic CFLMs, six with microcystic CFLMs, and 22 with mixed-type malformations. Mean post-intervention follow-up was 27.5 months. Fifty-eight patients underwent sclerotherapy (median: two treatments). Doxycycline and/or bleomycin were used in 95% of patients. After sclerotherapy, 97% of macrocystic CFLMs improved significantly compared to 82% of mixed and 67% of microcystic lesions. Sixteen children underwent surgical resection with 75% significantly improving; two additional patients were successfully treated with sclerotherapy after debulking surgery. Six children received sirolimus for microcystic disease, of which 33% significantly improved.ConclusionSclerotherapy is very effective for macrocystic components of CFLMs, albeit less so for microcystic disease. Microcystic CFLMs frequently require surgical resection. Sirolimus is a helpful therapeutic adjunct, particularly for microcystic lesions, but more study is needed.Level of EvidenceLevel II, prognosis study