IMPORTANCE Retinal telescreening for evaluation of diabetic retinopathy (DR) in the primary care setting may be useful in reaching rural and underserved patients.OBJECTIVES To evaluate telemedicine retinal screenings for patients with type 1 or 2 diabetes and identify factors for ophthalmology referral in the North Carolina Diabetic Retinopathy Telemedicine Network. DESIGN, SETTING, AND PARTICIPANTSA preimplementation and postimplementation evaluation was conducted from January 6, 2014, to November 1, 2015, at 5 primary care clinics serving rural and underserved populations in North Carolina among 1787 adult patients with type 1 or 2 diabetes who received primary care at the clinics and obtained retinal telescreening to determine the presence and severity of DR. A total of 1661 patients with complete data were included in the statistical analysis.INTERVENTION Nonmydriatic fundus photography with remote interpretation by an expert. MAIN OUTCOMES AND MEASURESNumber of patients recruited, level of detected DR, change in rates of screening, rate of ophthalmology referral, percentage of completed referrals, and patient characteristics associated with varying levels of DR. RESULTSOf the 1661 patients (1041 women and 620 men; mean [SD] age, 55.4 [12.7] years), 1323 patients (79.7%) had no DR, 183 patients (11.0%) had DR without a need for an ophthalmology referral, and 155 patients (9.3%) had DR with a need for an ophthalmology referral. The mean rate of screening for DR before implementation of the program was 25.6% (1512 of 5905), which increased to 40.4% (1884 of 4664) after implementation. A total of 93 referred patients (60.0%) completed an ophthalmology referral visit within the study period. Older patients (odds ratio [OR], 1.28; 95% CI, 1.11-1.48) and African American patients (OR, 1.84; 95% CI, 1.24-2.73) or other racial/ethnic minorities (OR, 2.19; 95% CI, 1.16-4.11) had greater odds of requiring an ophthalmology referral compared with white and/or younger patients. Patients with higher hemoglobin A 1c levels (OR, 1.19 per unit change; 95% CI, 1.13-1.25 per unit change) and longer duration of diabetes (OR, 1.76 per decade; 95% CI, 1.53-2.02 per decade) had greater odds of DR requiring an ophthalmology referral. History of stroke (OR, 1.65; 95% CI, 1.10-2.48) and kidney disease (OR, 1.59; 95% CI, 1.10-2.31) were strongly associated with DR and ophthalmology referral. CONCLUSIONS AND RELEVANCEWhen implemented in the primary care setting, retinal telescreening increased the rate of evaluation for DR for patients in rural and underserved settings. This strategy may also increase access to care for minorities and patients with DR requiring treatment.

We study the nonparametric least squares estimator (LSE) of a multivariate convex regression function. The LSE, given as the solution to a quadratic program with O(n 2 ) linear constraints (n being the sample size), is difficult to compute for large problems. Exploiting problem specific structure, we propose a scalable algorithmic framework based on the augmented Lagrangian method to compute the LSE. We develop a novel approach to obtain smooth convex approximations to the fitted (piecewise affine) convex LSE and provide formal bounds on the quality of approximation. When the number of samples is not too large compared to the dimension of the predictor, we propose a regularization scheme -Lipschitz convex regression -where we constrain the norm of the subgradients, and study the rates of convergence of the obtained LSE. Our algorithmic framework is simple and flexible and can be easily adapted to * 1 arXiv:1509.08165v1 [stat.CO] 28 Sep 2015 handle variants: estimation of a non-decreasing/non-increasing convex/concave (with or without a Lipschitz bound) function. We perform numerical studies illustrating the scalability of the proposed algorithm.

Objectives. To examine the characteristics and temporal trends of yoga, tai chi, and qigong (YTQ) use among US adults. Methods. Using the 2002, 2007, 2012, and 2017 National Health Interview Surveys, we examined the prevalence, patterns, and predicting factors of YTQ use by Taylor series linear regression, the Wald F χ2 test, and multivariable logistic regression models (n = 116 404). Results. YTQ use increased from 5.8% in 2002 to 14.5% in 2017 (P ≤ .001). Only 6.6% of YTQ users were referred by their medical doctors, and approximately one third disclosed their use of YTQ to medical professionals. Reasons for using YTQ included (1) YTQ was beneficial, (2) YTQ focused on the whole person, and (3) YTQ was natural. Acute and chronic pain, arthritis, and depression were the top 3 medical conditions for which people used YTQ the most. Conclusions. YTQ use is increasing substantially, mainly because of its natural and holistic healing approach toward health and chronic diseases. Future studies aiming to explore how to best integrate YTQ into the current health care system are warranted.