Impact of Different Mammography Systems on Artificial Intelligence Performance in Breast Cancer Screening.

Artificial intelligence (AI) tools may assist breast screening mammography programs, but limited evidence supports their generalizability to new settings. This retrospective study used a 3-year dataset (April 1, 2016-March 31, 2019) from a U.K. regional screening program. The performance of a commercially available breast screening AI algorithm was assessed with a prespecified and site-specific decision threshold to evaluate whether its performance was transferable to a new clinical site. The dataset consisted of women (aged approximately 50-70 years) who attended routine screening, excluding self-referrals, those with complex physical requirements, those who had undergone a previous mastectomy, and those who underwent screening that had technical recalls or did not have the four standard image views. In total, 55 916 screening attendees (mean age, 60 years ± 6 [SD]) met the inclusion criteria. The prespecified threshold resulted in high recall rates (48.3%, 21 929 of 45 444), which reduced to 13.0% (5896 of 45 444) following threshold calibration, closer to the observed service level (5.0%, 2774 of 55 916). Recall rates also increased approximately threefold following a software upgrade on the mammography equipment, requiring per-software version thresholds. Using software-specific thresholds, the AI algorithm would have recalled 277 of 303 (91.4%) screen-detected cancers and 47 of 138 (34.1%) interval cancers. AI performance and thresholds should be validated for new clinical settings before deployment, while quality assurance systems should monitor AI performance for consistency. Keywords: Breast, Screening, Mammography, Computer Applications-Detection/Diagnosis, Neoplasms-Primary, Technology Assessment Supplemental material is available for this article. © RSNA, 2023.

Postoperative Compliance and Return to Work After Rotator Cuff Repair: Value of an Interactive Online Rehabilitation Program Among Patients Treated Under Workers' Compensation.

Outcomes after rotator cuff repair (RCR) typically are poorer for workers' compensation (WC) than for patients with private insurance. This study examined augmentation of the traditional rehabilitation protocol with an online exercise program. Between March 2016 and July 2018, 48 WC patients who underwent RCR were introduced to a digital rehabilitation program (application). Patients were divided into patients who used the application along with traditional physical therapy (PT) (group 1) and patients who underwent only traditional PT (group 2). Patient performance was assessed using standardized patient-reported outcome measures (PROMs), return-to-work (RTW) status, number of PT sessions, and complication/revision rate. Patients were monitored for a minimum of 1 year postoperatively. A significant improvement in RTW period was noted for group 1 patients, who resumed modified duty 10 weeks earlier than group 2 patients. A positive trend also was noted for return to full duty, with group 1 resuming full duty 7 weeks before group 2. There were no complications or reoperations in group 1 compared with 4 (16%) complications in group 2. The 2 groups underwent the same average number of PT sessions (27 sessions). Patient-reported outcome measures were captured only for group 1, which demonstrated average postoperative improvement of 3 points on a visual analog scale and 32.5 points in American Shoulder and Elbow Surgeons scores. Incorporating an online exercise program within the traditional rehabilitation protocol for WC patients undergoing RCR resulted in earlier return to work and was associated with better pain relief, greater return of function, and lower complication rate. [Orthopedics. 2021;44(2):e197-e202.].

Preoperative physiotherapy is cost-effective for preventing pulmonary complications after major abdominal surgery: a health economic analysis of a multicentre randomised trial.

QUESTION: Is preoperative physiotherapy cost-effective in reducing postoperative pulmonary complications (PPC) and improving quality-adjusted life years (QALYs) after major abdominal surgery? DESIGN: Cost-effectiveness analysis from the hospitals' perspective within a multicentre randomised controlled trial with concealed allocation, blinded assessors and intention-to-treat analysis. PARTICIPANTS: Four hundred and forty-one adults awaiting elective upper abdominal surgery attending pre-anaesthetic clinics at three public hospitals in Australia and New Zealand. INTERVENTIONS: The experimental group received an information booklet and a 30-minute face-to-face session, involving respiratory education and breathing exercise training, with a physiotherapist. The control group received the information booklet only. OUTCOME MEASURES: The probability of cost-effectiveness and incremental net benefits was estimated using bootstrapped incremental PPC and QALY cost-effectiveness ratios plotted on cost-effectiveness planes and associated probability curves through a range of willingness-to-pay amounts. Cost-effectiveness modelling utilised 21-day postoperative hospital cost audit data and QALYs estimated from Short Form-Six Domain health utilities and mortality to 12 months. RESULTS: Preoperative physiotherapy had 95% probability of being cost-effective with an incremental net benefit to participating hospitals of A$4,958 (95% CI 10 to 9,197) for each PPC prevented, given that the hospitals were willing to pay $45,000 to provide the service. Cost-utility for QALY gains was less certain. Sensitivity analyses strengthened cost-effectiveness findings. Improved cost-effectiveness and QALY gains were detected when experienced physiotherapists delivered the intervention. CONCLUSIONS: Preoperative physiotherapy aimed at preventing PPCs was highly likely to be cost-effective from the hospitals' perspective. For each PPC prevented, preoperative physiotherapy is likely to cost the hospitals less than the costs estimated to treat a PPC after surgery. Potential QALY gains require confirmation. TRIAL REGISTRATION: ACTRN12613000664741.

The MOSAICC study: Assessing feasibility for biological sample collection in epidemiology studies and comparison of DNA yields from saliva and whole blood samples.

Biological sample collection is becoming more common in epidemiology research to obtain DNA for genetic analysis. There are many different DNA collection methods but little evidence on their relative effectiveness. Therefore, we took the opportunity of a prospective case-control study in myeloproliferative neoplasms (MPNs) to compare DNA yield from 8.5 mL PAXgene tubes for whole blood collection versus 2 mL Oragene OG-500 saliva collection kits. MPNs include polycythaemia vera, essential thrombocythaemia, and primary myelofibrosis. These are rare diseases and our exploratory case-control study (MOSAICC) sought to improve knowledge regarding their aetiology and to determine optimal methodology for a larger UK-wide study. Overall, 233 participants were recruited to the MOSIACC study, and we collected 187 blood and 214 saliva samples. The mean DNA yield from blood was 659.18 ng/µL, significantly higher than the mean DNA yield from saliva samples (275.79 ng/µL). The higher provision of saliva samples might reflect its non-invasive and more convenient nature, compared to blood sample provision. The differences in mean DNA yields might reflect differences in clinical assistance, adherence to instructions, and health status of individuals. In conclusion, both sample collection techniques are simple, effective, and suitable for DNA collection for genetic analysis in future epidemiological research studies but OG-500 kits offer a less invasive alternative for those who refuse to provide blood.

Knowledge or noise? Making sense of General Practitioners' and Consultant use of 2-week-wait referrals for suspected cancer.

BACKGROUND: Early diagnosis and treatment of cancer is the goal of the 2-week-wait referral pathway (2WW). Variation exists between General Practice use of 2WW and rates of consultant reprioritisation of GP referral from routine to 2WW (Consultant Upgrade). We investigated variation in General Practice and Consultant Upgrade 2WW referral activity. METHODS: Data from 185 000 referrals and 29 000 cancers recorded between 2011 and 2013 from the Northern Ireland Cancer Waiting Time database (CaPPS) were analysed to ascertain standardised referral rate ratios, detection rate (DR) (=sensitivity) and conversion rate (CR) (=positive predictive value) for Practice 2WW referrals and Consultant Upgrade 2WW. Metrics were compared using Spearman's rank correlation co-efficients. RESULTS: There was consistency in Practice and Consultant Upgrade 2WW referral rates over time, though not for annual DR (Spearman's ρ<0.37) or CR (Spearman's ρ<0.26). Practice 2WW referral rates correlated negatively with CR and positively with DR while correlations between DR and CR were restricted to single-year comparisons in Practice 2WW. In Consultant Upgrade, 2WW CR and DR were strongly correlated but only when the same cancers were included in both rates. CONCLUSIONS: Results suggest 'random case mix' explains previously reported associations between CR and DR with more 'hard to detect' cancers in some Practices than in others in a given year corresponding to lower DR and CR. Use of Practice and Consultant Upgrade 2WW referral metrics to gauge General Practice performance may be misleading.

Myeloproliferative neoplasm patient symptom burden and quality of life: evidence of significant impairment compared to controls.

The myeloproliferative neoplasms (MPN) including polycythaemia vera (PV), essential thrombocythaemia and primary myelofibrosis (PMF) are rare diseases contributing to significant morbidity. Symptom management is a prime treatment objective but current symptom assessment tools have not been validated compared to the general population. The MPN-symptom assessment form (MPN-SAF), a reliable and validated clinical tool to assess MPN symptom burden, was administered to MPN patients (n = 106) and, for the first time, population controls (n = 124) as part of a UK case-control study. Mean symptom scores were compared between patients and controls adjusting for potential confounders. Mean patient scores were compared to data collected by the Mayo Clinic, USA on 1,446 international MPN patients to determine patient group representativeness. MPN patients had significantly higher mean scores than controls for 25 of the 26 symptoms measured (P < 0.05); fatigue was the most common symptom (92.4% and 78.1%, respectively). Female MPN patients suffered worse symptom burden than male patients (P < 0.001) and substantially worse burden than female controls (P < 0.001). Compared to the Mayo clinic patients, MPN-UK patients reported similar symptom burden but lower satiety (P = 0.046). Patients with PMF reported the worst symptom burden (88.3%); significantly higher than PV patients (P < 0.001). For the first time we report quality of life was worse in MPN-UK patients compared with controls (P < 0.001).

Peripheral arterial tone: assessment of microcirculatory function in pregnancy.

OBJECTIVE: Endothelial dysfunction is known to play a key role in the pathogenesis of preeclampsia, but the majority of methods for its detection are too invasive to be used in pregnancy. In this study we report a novel method - peripheral arterial tonometry (PAT) - for examining microcirculatory function in pregnancy. METHODS: One hundred and eighty women with at least two risk factors for preeclampsia were examined at gestational weeks 16 and 28; 80 women were examined at 6-9 months postnatally. Twenty-four women developed preeclampsia or pregnancy-induced hypertension (cases), 156 remained normotensive (controls). PAT was measured using fingertip pneumatic probes; after baseline recordings the study arm was occluded with a blood pressure cuff then released after 5 min, causing reactive hyperaemia. PAT recordings pre and post occlusion were used to generate the reactive hyperaemia index (RHI). RESULTS: RHI was significantly lower at gestational week 28 compared to week 16, both in cases and controls. Baseline pulse amplitude was significantly higher at week 28 compared to week 16. There was no difference in RHI at either week 16 or 28 between cases and controls. Postnatally, there was no difference in RHI between cases and controls, but baseline pulse amplitude was lower in affected women. CONCLUSION: PAT and other methods which rely on flow-mediated dilatation for detection of endothelial dysfunction are less likely to be reliable in later pregnancy, when women are more vasodilated. PAT did not predict the development of hypertensive pregnancy complications, but demonstrated a relative peripheral vasoconstriction in affected women postnatally.

Socio-economic and other correlates of Kaposi sarcoma-associated herpesvirus seroprevalence among older adults in Sicily.

The virus that causes Kaposi sarcoma, KS-associated herpesvirus (KSHV, also known as human herpesvirus 8) has an unusual distribution and poorly characterized modes of transmission. To clarify these issues, socio-demographic correlates of KSHV seroprevalence were examined in a population-based study. In 1,154 randomly sampled adults (aged 32- 92, mean 71 years) throughout Sicily, KSHV antibodies were detected with four assays and a conservative algorithm. Seroprevalence was re-weighted to the population. Odds ratios with 95% confidence intervals (OR, CI) from multivariate logistic regression were used to estimate associations of seroprevalence with interview data. KSHV seroprevalence was 8.5%, including 5.3% among men (N = 848) and 11.5% among women (N = 306, P = 0.22). Seroprevalence was higher with residence in a smaller community during childhood (P(trend) = 0.03) and working with plants/soil during adulthood (OR 2.9, CI 1.1-7.9); these were especially strong among women. Among men, seroprevalence was significantly associated with lower education (OR 2.6, CI 1.1-5.9) and migration to a larger community (OR 0.3, CI 0.1-0.9). Other demographic and household variables were unrelated to seroprevalence. From these data, KSHV in Sicily appears to be related to low socio-economic status, but micro-endemicity in small communities cannot be excluded.