• Association Between Hormone-Modulating Breast Cancer Therapies and Incidence of Neurodegenerative Outcomes for Women With Breast Cancer

      Branigan, Gregory L; Soto, Maira; Neumayer, Leigh; Rodgers, Kathleen; Brinton, Roberta Diaz; Univ Arizona, Ctr Innovat Brain Sci; Univ Arizona, Coll Med, Dept Pharmacol; Univ Arizona, MD PhD Training Program, Coll Med; Univ Arizona, Dept Surg, Coll Med; Univ Arizona, Dept Obstet & Gynecol, Coll Med; et al. (AMER MEDICAL ASSOC, 2020-03-02)
      Question Is hormone-modulating therapy associated with neurodegenerative disease in women with breast cancer? Findings In this cohort study of 57 & x202f;843 perimenopausal- to postmenopausal-aged women with breast cancer, exposure to hormone-modulating therapy (tamoxifen and aromatase inhibitors, especially exemestane) was associated with a significant decrease in the number of women who received a diagnosis of neurodegenerative disease, most specifically Alzheimer disease. Meaning With the increased life expectancy seen after treatment, therapy selection for breast cancer should include a careful discussion of the risks and benefits of each treatment option that may be associated with a reduced risk of neurodegenerative disease. Importance The association between exposure to hormone-modulating therapy (HMT) as breast cancer treatment and neurodegenerative disease (NDD) is unclear. Objective To determine whether HMT exposure is associated with the risk of NDD in women with breast cancer. Design, Setting, and Participants This retrospective cohort study used the Humana claims data set from January 1, 2007, to March 31, 2017. The Humana data set contains claims from private-payer and Medicare insurance data sets from across the United States with a population primarily residing in the Southeast. Patient claims records were surveyed for a diagnosis of NDD starting 1 year after breast cancer diagnosis for the duration of enrollment in the claims database. Participants were 57 & x202f;843 women aged 45 years or older with a diagnosis of breast cancer. Patients were required to be actively enrolled in Humana claims records for 6 months prior to and at least 3 years after the diagnosis of breast cancer. The analyses were conducted between January 1 and 15, 2020. Exposure Hormone-modulating therapy (selective estrogen receptor modulators, estrogen receptor antagonists, and aromatase inhibitors). Main Outcomes and Measures Patients receiving HMT for breast cancer treatment were identified. Survival analysis was used to determine the association between HMT exposure and diagnosis of NDD. A propensity score approach was used to minimize measured and unmeasured selection bias. Results Of the 326 & x202f;485 women with breast cancer in the Humana data set between 2007 and 2017, 57 & x202f;843 met the study criteria. Of these, 18 & x202f;126 (31.3%; mean [SD] age, 76.2 [7.0] years) received HMT, whereas 39 & x202f;717 (68.7%; mean [SD] age, 76.8 [7.0] years) did not receive HMT. Mean (SD) follow-up was 5.5 (1.8) years. In the propensity score-matched population, exposure to HMT was associated with a decrease in the number of women who received a diagnosis of NDD (2229 of 17 878 [12.5%] vs 2559 of 17 878 [14.3%]; relative risk, 0.89; 95% CI, 0.84-0.93; P < .001), Alzheimer disease (877 of 17 878 [4.9%] vs 1068 of 17 878 [6.0%]; relative risk, 0.82; 95% CI, 0.75-0.90; P < .001), and dementia (1862 of 17 878 [10.4%] vs 2116 of 17 878 [11.8%]; relative risk, 0.88; 95% CI, 0.83-0.93; P < .001). The number needed to treat was 62.51 for all NDDs, 93.61 for Alzheimer disease, and 69.56 for dementia. Conclusions and Relevance Among patients with breast cancer, tamoxifen and steroidal aromatase inhibitors were associated with a decrease in the number who received a diagnosis of NDD, specifically Alzheimer disease and dementia. This cohort study uses the Humana claims data set to examine whether exposure to hormone-modulating therapy is associated with the risk of neurodegenerative disease in women with breast cancer.
    • Association Between Noninvasive Fibrosis Markers and Postoperative Mortality After Hepatectomy for Hepatocellular Carcinoma

      Maegawa, Felipe B; Shehorn, Lauren; Aziz, Hassan; Kettelle, John; Jie, Tun; Riall, Taylor S; Univ Arizona, Dept Surg (AMER MEDICAL ASSOC, 2019-01-18)
      IMPORTANCE The selection criteria for hepatectomy for hepatocellular carcinoma (HCC) is not well established. The role of noninvasive fibrosis markers in this setting is unknown in the US population. OBJECTIVE To evaluate whether aspartate aminotransferase-platelet ratio index (APRI) and fibrosis 4 (Fib4) values are associated with perioperative mortality and overall survival after hepatectomy for HCC. DESIGN, SETTING, AND PARTICIPANTS In a multicenter cohort study, Veterans Administration Corporate Data Warehouse was used to evaluate a retrospective cohort of 475 veterans who underwent hepatectomy for HCC between January 1, 2000, and December 31, 2012, in Veterans Administration hospitals. Data analysis occurred between September 30, 2016, and December 30, 2017. Logistic regression, survival analysis, and change in concordance index analysis were performed to evaluate the association between APRI and Fib4 values and mortality. EXPOSURES The cohort was stratified based on preoperative APRI and Fib4 values. Analysis was performed accounting for the validated and established predictors of outcome. MAIN OUTCOMES AND MEASURES Thirty-day mortality, 90-day mortality, and overall survival were the primary outcomes. An APRI value greater than 1.5 was considered high risk (cirrhosis), and an Fib4 value greater than 4.0 was considered high risk (advanced fibrosis). Portal hypertension (diagnosis of ascites or encephalopathy indicates presence of portal hypertension) and Child-Turcotte-Pugh (CTP) class (A indicates preserved liver function; B, mild to moderate liver dysfunction) served as 2 other measures of liver function. RESULTS A total of 475 patients with HCC underwent hepatectomy. The mean (SD) age was 65.6 (9.4) years; Model for End-Stage Liver Disease score, 8.9 (3.1); and body mass index, 28.1 (4.9) (calculated as weight in kilograms divided by height in meters squared). A total of 361 patients (76.0%) were men, 294 (61.9%) were white; 308 (64.8%) were hepatitis C positive, and 346 (72.8%) were categorized as CTP class A. The most common surgical procedure was partial lobectomy, with 321 (67.6%) procedures. The APRI value greater than 1.5 vs 1.5 or lower was associated with increased 30-day mortality (odds ratio [OR], 6.45; 95% CI, 2.80-14.80) and 90-day mortality (OR, 2.65; 95% CI, 1.35-5.22), as was Fib4 greater than 4.0 vs Fib4 4.0 or lower for 30-day mortality (OR, 5.41; 95% CI, 2.35-12.50) and 90-day mortality (OR, 2.74; 95% CI, 1.41-5.35). Survival analysis showed that overall survival was significantly different for APRI greater than 1.5 vs 1.5 or lower (mean survival time, 3.6 vs 5.4 years; log-rank P <.001) and Fib4 greater than 4.0 vs 4.0 or lower (mean survival time, 4.1 vs 5.3 years; log rank P =.01). Adjusted Cox proportional hazards regression analysis revealed that elevated APRI was significantly associated with worse survival (hazard ratio [HR], 1.13; 95% CI, 1.03-1.23) but Fib4 values were not (HR, 1.04; 95% CI, 0.99-1.09). Change in concordance index showed that APRI and Fib4 improved the ability of CTP class and portal hypertension to predict postoperative mortality. CONCLUSIONS AND RELEVANCE Elevated APRI and Fib4 values, which are noninvasive markers of fibrosis, were associated with higher perioperative mortality. The APRI was also associated with worse overall survival. Use of APRI and Fib4 measures improved the ability of established markers to predict postoperative mortality. These findings suggest incorporating APRI and Fib4 to the selection process for hepatectomy for HCC as predictors associated with mortalitymay be warranted.
    • Association Between Thyroid Disorders and Colorectal Cancer Risk in Adult Patients in Taiwan

      L’Heureux, Abby; Wieland, Daniel R.; Weng, Chien-Huan; Chen, Yi-Huei; Lin, Ching-Heng; Lin, Tseng-Hsi; Weng, Chien-Hsiang; Univ Arizona (AMER MEDICAL ASSOC, 2019-05-17)
      IMPORTANCE Thyroid hormones have been shown to affect several important pathways in cancer development, including colorectal cancer (CRC). Clinical studies examining the association between thyroid disorders and colorectal cancer have conflicting results and have predominantly involved white populations. OBJECTIVE To determine if a diagnosis of hyperthyroidism or hypothyroidism is associated with the risk of developing colorectal cancer in an East Asian population. DESIGN, SETTING, AND PARTICIPANTS This nationwide population-based case-control study was conducted from April 27, 2018, to November 8, 2018, using the Taiwanese National Health Insurance Research Database. Participants were adults (n = 139 426) either with a new diagnosis (between 2008 and 2013) of primary colorectal cancer without a history of cancer, or without cancer. Cases and controls were matched 11 by age, sex, and index date. Diagnosis of hyperthyroidism or hypothyroidism prior to the diagnosis of colorectal cancer (or the same index date in controls) was then determined. MAIN OUTCOMES AND MEASURES Risk differences in developing colorectal cancer among patients with a medical history of hyperthyroidism or hypothyroidism. RESULTS A total of 139 426 patients were included in the study, and 69 713 individuals made up each case and control group, which were both predominantly male (39 872 [57.2%]). The mean (SD) age for those with CRC was 65.8 (13.7) years and for those without CRC was 66.0 (13.6) years. Both hyperthyroidism (adjusted odds ratio [aOR], 0.77; 95% CI, 0.69-0.86; P < .001) and hypothyroidism (aOR, 0.78; 95% CI, 0.65-0.94; P = .008) were associated with a decreased risk of being diagnosed with colorectal cancer. An inverse association of rectal cancer was found among patients aged 50 years or older with a history of hypothyroidism despite treatment (aOR, 0.54; 95% CI, 0.39-0.74; P < .001). A history of hyperthyroidism in all age groups was associated with a lower risk of colon cancer (aOR, 0.74; 95% CI, 0.64-0.85; P < .001), with a stronger association seen among those younger than 50 years (aOR, 0.55; 95% CI, 0.36-0.85; P = .007). CONCLUSIONS AND RELEVANCE In this study, hypothyroidism appeared to be associated with a lower risk of rectal cancer, whereas hyperthyroidism appeared to be associated with a lower risk of colon cancer. Because of this, biochemical in vivo research and epidemiologic studies appear to be needed to further clarify the nature of these associations.
    • Association of Light Physical Activity Measured by Accelerometry and Incidence of Coronary Heart Disease and Cardiovascular Disease in Older Women

      LaCroix, Andrea Z; Bellettiere, John; Rillamas-Sun, Eileen; Di, Chongzhi; Evenson, Kelly R; Lewis, Cora E; Buchner, David M; Stefanick, Marcia L; Lee, I-Min; Rosenberg, Dori E; et al. (AMER MEDICAL ASSOC, 2019-03-15)
      IMPORTANCE To our knowledge, no studies have examined light physical activity (PA) measured by accelerometry and heart disease in older women. OBJECTIVE To investigate whether higher levels of light PA were associated with reduced risks of coronary heart disease (CHD) or cardiovascular disease (CVD) in older women. DESIGN, SETTING, AND PARTICIPANTS Prospective cohort study of older women from baseline (March 2012 to April 2014) through February 28, 2017, for up to 4.91 years. The setting was community-dwelling participants from the Women's Health Initiative. Participants were ambulatory women with no history of myocardial infarction or stroke. EXPOSURES Data from accelerometers worn for a requested 7 days were used to measure light PA. MAIN OUTCOMES AND MEASURES Cox proportional hazards regression models estimated hazard ratios (HRs) and 95% CIs for physician-adjudicated CHD and CVD events across light PA quartiles adjusting for possible confounders. Light PA was also analyzed as a continuous variable with and without adjustment for moderate to vigorous PA (MVPA). RESULTS Among 5861 women (mean [SD] age, 78.5 [6.7] years), 143 CHD events and 570 CVD events were observed. The HRs for CHD in the highest vs lowest quartiles of light PA were 0.42 (95% CI, 0.25-0.70; P for trend <. 001) adjusted for age and race/ethnicity and 0.58 (95% CI, 0.34-0.99; P for trend = .004) after additional adjustment for education, current smoking, alcohol consumption, physical functioning, comorbidity, and self-rated health. Corresponding HRs for CVD in the highest vs lowest quartiles of light PA were 0.63 (95% CI, 0.49-0.81; P for trend <. 001) and 0.78 (95% CI, 0.60-1.00; P for trend = .004). The HRs for a 1-hour/day increment in light PA after additional adjustment for MVPA were 0.86 (95% CI, 0.73-1.00; P for trend = .05) for CHD and 0.92 (95% CI, 0.85-0.99; P for trend = .03) for CVD. CONCLUSIONS AND RELEVANCE The present findings support the conclusion that all movement counts for the prevention of CHD and CVD in older women. Large, pragmatic randomized trials are needed to test whether increasing light PA among older women reduces cardiovascular risk.
    • Association of Physical Activity and Fracture Risk Among Postmenopausal Women

      LaMonte, Michael J; Wactawski-Wende, Jean; Larson, Joseph C; Mai, Xiaodan; Robbins, John A; LeBoff, Meryl S; Chen, Zhao; Jackson, Rebecca D; LaCroix, Andrea Z; Ockene, Judith K; et al. (AMER MEDICAL ASSOC, 2019-10-25)
      IMPORTANCE Physical activity is inversely associated with hip fracture risk in older women. However, the association of physical activity with fracture at other sites and the role of sedentary behavior remain unclear. OBJECTIVE To assess the associations of physical activity and sedentary behavior with fracture incidence among postmenopausal women. DESIGN, SETTING, AND PARTICIPANTS The Women's Health Initiative prospective cohort study enrolled 77 206 postmenopausal women aged 50 to 79 years between October 1993 and December 1998 at 40 US clinical centers. Participants were observed for outcomes through September 2015, with data analysis conducted from June 2017 to August 2019. EXPOSURES Self-reported physical activity and sedentary time. MAIN OUTCOMES AND MEASURES Hazard ratios (HRs) and 95% CIs for total and site-specific fracture incidence. RESULTS During a mean (SD) follow-up period of 14.0 (5.2) years among 77 206 women (mean [SD] age, 63.4 [7.3] years; 66 072 [85.6%] white), 25 516 (33.1%) reported a first incident fracture. Total physical activity was inversely associated with the multivariable-adjusted risk of hip fracture (>17.7 metabolic equivalent [MET] h/wk vs none: HR, 0.82; 95% CI, 0.72-0.95; P for trend < .001). Inverse associations with hip fracture were also observed for walking (>7.5 MET h/wk vs none: HR, 0.88; 95% CI, 0.78-0.98; P for trend = .01), mild activity (HR, 0.82; 95% CI, 0.73-0.93; P for trend = .003), moderate to vigorous activity (HR, 0.88; 95% CI, 0.81-0.96; P for trend = .002), and yard work (HR, 0.90; 95% CI, 0.82-0.99; P for trend = .04). Total activity was positively associated with knee fracture (>17.7 MET h/wk vs none: HR, 1.26; 95% CI, 1.05-1.50; P for trend = .08). Mild activity was associated with lower risks of clinical vertebral fracture (HR, 0.87; 95% CI, 0.78-0.96; P for trend = .006) and total fractures (HR, 0.91; 95% CI, 0.87-0.94; P for trend < .001). Moderate to vigorous activity was positively associated with wrist or forearm fracture (HR, 1.09; 95% CI, 1.03-1.15; P for trend = .004). After controlling for covariates and total physical activity, sedentary time was positively associated with total fracture risk (>9.5 h/d vs <6.5 h/d: HR, 1.04; 95% CI, 1.01-1.07; P for trend = .01). When analyzed jointly, higher total activity mitigated some of the total fracture risk associated with sedentary behavior. Analysis of time-varying exposures resulted in somewhat stronger associations for total physical activity, whereas those for sedentary time were materially unchanged. CONCLUSIONS AND RELEVANCE In older ambulatory women, higher total physical activity was associated with lower total and hip fracture risk but higher knee fracture risk. Mild activity and walking were associated with lower hip fracture risk, a finding with important public health implications because these activities are common in older adults. The positive association between sedentary time and total fracture risk requires further investigation.
    • Effect of a Positive Psychological Intervention on Pain and Functional Difficulty Among Adults With Osteoarthritis

      Hausmann, Leslie R. M.; Youk, Ada; Kwoh, C. Kent; Gallagher, Rollin M.; Weiner, Debra K.; Vina, Ernest R.; Obrosky, D. Scott; Mauro, Genna T.; McInnes, Shauna; Ibrahim, Said A.; et al. (AMER MEDICAL ASSOC, 2018-09)
      IMPORTANCE Positive psychological interventions for improving health have received increasing attention recently. Evidence on the impact of such interventions on pain, and racial disparities in pain, is limited. OBJECTIVE To assess the effects of a positive psychological intervention on pain and functional difficulty in veterans with knee osteoarthritis. DESIGN, SETTING, AND PARTICIPANTS The Staying Positive With Arthritis Study is a large, double-blinded randomized clinical trial powered to detect race differences in self-reported pain in response to a positive psychological intervention compared with a neutral control intervention. Data were collected from 2 urban Veterans Affairs medical centers. Participants included non-Hispanic white and non-Hispanic African American patients aged 50 years or older with a diagnosis of osteoarthritis. Mailings were sent to 5111 patients meeting these criteria, of whom 839 were fully screened, 488 were eligible, and 360 were randomized. Enrollment lasted from July 8, 2015, to February 1, 2017, with follow-up through September 6, 2017. INTERVENTIONS The intervention comprised a 6-week series of evidence-based activities to build positive psychological skills (eg, gratitude and kindness). The control program comprised similarly structured neutral activities. Programs were delivered via workbook and weekly telephone calls with interventionists. MAIN OUTCOMES AND MEASURES The primary outcomes were self-reported pain and functional difficulty measured using the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC; range 0-100). Secondary outcomes included affect balance and life satisfaction. RESULTS The sample included 180 non-Hispanic white patients and 180 non-Hispanic African American patients (mean [SD] age, 64.2 [8.8] years; 76.4% were male). Mean (SD) baseline scores for WOMAC pain and functional difficulty were 48.8 (17.6) and 46.8 (18.1). respectively. Although both decreased significantly over time (pain: chi(2)(3) = 49.50. P < .001; functional difficulty: chi(2)(3) = 22.11, P < .001), differences were small and did not vary by treatment group or race. Exploratory analyses suggested that the intervention had counter-intuitive effects on secondary outcomes. CONCLUSIONS AND RELEVANCE The results of this randomized clinical trial do not support the use of positive psychological interventions as a stand-alone treatment for pain among white or African American veterans with knee osteoarthritis. Adaptations are needed to identify intervention components that resonate with this population, and the additive effect of incorporating positive psychological interventions into more comprehensive pain treatment regimens should be considered.
    • Effect of a Virtual Patient Navigation Program on Behavioral Health Admissions in the Emergency Department: A Randomized Clinical Trial

      Roberge, Jason; McWilliams, Andrew; Zhao, Jing; Anderson, William E; Hetherington, Timothy; Zazzaro, Christine; Hardin, Elisabeth; Barrett, Amy; Castro, Manuel; Balfour, Margaret E; et al. (AMER MEDICAL ASSOC, 2020-01-29)
      This randomized clinical trial assesses whether the availability of a 45-day behavioral health-virtual patient navigation program decreases hospitalization among adult patients presenting to the emergency department with a behavioral health crisis or need. Importance The number of patients presenting to emergency departments (EDs) for psychiatric care continues to increase. Psychiatrists often make a conservative recommendation to admit patients because robust outpatient services for close follow-up are lacking. Objective To assess whether the availability of a 45-day behavioral health-virtual patient navigation program decreases hospitalization among patients presenting to the ED with a behavioral health crisis or need. Design, Setting, and Participants This randomized clinical trial enrolled 637 patients who presented to 6 EDs spanning urban and suburban locations within a large integrated health care system in North Carolina from June 12, 2017, through February 14, 2018; patients were followed up for up to 45 days. Eligible patients were aged 18 years or older, with a behavioral health crisis and a completed telepsychiatric ED consultation. The availability of the behavioral health-virtual patient navigation intervention was randomly allocated to specific days (Monday through Friday from 7 am to 7 pm) so that, in a 2-week block, there were 5 intervention days and 5 usual care days; 323 patients presented on days when the program was offered, and 314 presented on usual care days. Data analysis was performed from March 7 through June 13, 2018, using an intention-to-treat approach. Interventions The behavioral health-virtual patient navigation program included video contact with a patient while in the ED and telephonic outreach 24 to 72 hours after discharge and then at least weekly for up to 45 days. Main Outcomes and Measures The primary outcome was the conversion of an ED encounter to hospital admission. Secondary outcomes included 45-day follow-up encounters with a self-harm diagnosis and postdischarge acute care use. Results Among 637 participants, 358 (56.2%) were men, and the mean (SD) age was 39.7 (16.6) years. The conversion rates were 55.1% (178 of 323) in the intervention group vs 63.1% (198 of 314) in the usual care group (odds ratio, 0.74; 95% CI, 0.54-1.02; P = .06). The percentage of patient encounters with follow-up encounters having a self-harm diagnosis was significantly lower in the intervention group compared with the usual care group (36.8% [119 of 323] vs 45.5% [143 of 314]; P = .03). Conclusions and Relevance Although the primary result did not reach statistical significance, there is a strong signal of potential positive benefit in an area that lacks evidence, suggesting that there should be additional investment and inquiry into virtual behavioral health programs. Question Does offering virtual patient navigation reduce admission rates for patients presenting to the emergency department with a behavioral health crisis? Findings In this randomized clinical trial, there were fewer admissions on days when the navigation program was available (55.1%) vs on days with usual care (63.1%), although the difference was not statistically significant. Significantly fewer patients who used the navigation program had a follow-up encounter involving a self-harm diagnosis within 45 days compared with patients who received usual care (36.8% vs 45.5%). Meaning Although the primary result did not reach statistical significance, there is a strong signal of potential positive benefit in an area that lacks evidence, suggesting that there should be additional investment and inquiry into this area.
    • Evaluation of Machine-Learning Algorithms for Predicting Opioid Overdose Risk Among Medicare Beneficiaries With Opioid Prescriptions

      Lo-Ciganic, Wei-Hsuan; Huang, James L; Zhang, Hao H; Weiss, Jeremy C; Wu, Yonghui; Kwoh, C Kent; Donohue, Julie M; Cochran, Gerald; Gordon, Adam J; Malone, Daniel C; et al. (AMER MEDICAL ASSOC, 2019-03-22)
      IMPORTANCE Current approaches to identifying individuals at high risk for opioid overdose target many patients who are not truly at high risk. OBJECTIVE To develop and validate a machine-learning algorithm to predict opioid overdose risk among Medicare beneficiaries with at least 1 opioid prescription. DESIGN, SETTING, AND PARTICIPANTS A prognostic study was conducted between September 1, 2017, and December 31, 2018. Participants (n = 560 057) included fee-for-service Medicare beneficiaries without cancer who filled 1 or more opioid prescriptions from January 1, 2011, to December 31, 2015. Beneficiaries were randomly and equally divided into training, testing, and validation samples. EXPOSURES Potential predictors (n = 268), including sociodemographics, health status, patterns of opioid use, and practitioner-level and regional-level factors, were measured in 3-month windows, starting 3 months before initiating opioids until loss of follow-up or the end of observation. MAIN OUTCOMES AND MEASURES Opioid overdose episodes from inpatient and emergency department claims were identified. Multivariate logistic regression (MLR), least absolute shrinkage and selection operator-type regression (LASSO), random forest (RF), gradient boosting machine (GBM), and deep neural network (DNN) were applied to predict overdose risk in the subsequent 3 months after initiation of treatment with prescription opioids. Prediction performance was assessed using the C statistic and other metrics (eg, sensitivity, specificity, and number needed to evaluate [NNE] to identify one overdose). The Youden index was used to identify the optimized threshold of predicted score that balanced sensitivity and specificity. RESULTS Beneficiaries in the training (n = 186 686), testing (n = 186 685), and validation (n = 186 686) samples had similar characteristics (mean [SD] age of 68.0 [14.5] years, and approximately 63% were female, 82% were white, 35% had disabilities, 41% were dual eligible, and 0.60% had at least 1 overdose episode). In the validation sample, the DNN (C statistic = 0.91; 95% CI, 0.88-0.93) and GBM (C statistic = 0.90; 95% CI, 0.87-0.94) algorithms outperformed the LASSO (C statistic = 0.84; 95% CI, 0.80-0.89), RF (C statistic = 0.80; 95% CI, 0.75-0.84), and MLR (C statistic = 0.75; 95% CI, 0.69-0.80) methods for predicting opioid overdose. At the optimized sensitivity and specificity, DNN had a sensitivity of 92.3%, specificity of 75.7%, NNE of 542, positive predictive value of 0.18%, and negative predictive value of 99.9%. The DNN classified patients into low-risk (76.2%[142 180] of the cohort), medium-risk (18.6%[34 579] of the cohort), and high-risk (5.2%[9747] of the cohort) subgroups, with only 1 in 10 000 in the low-risk subgroup having an overdose episode. More than 90% of overdose episodes occurred in the high-risk and medium-risk subgroups, although positive predictive values were low, given the rare overdose outcome. CONCLUSIONS AND RELEVANCE Machine-learning algorithms appear to perform well for risk prediction and stratification of opioid overdose, especially in identifying low-risk subgroups that have minimal risk of overdose.
    • Magnitude of and Characteristics Associated With the Treatment of Calcium Channel Blocker-Induced Lower-Extremity Edema With Loop Diuretics

      Vouri, Scott Martin; Jiang, Xinyi; Manini, Todd M; Solberg, Laurence M; Pepine, Carl; Malone, Daniel C; Winterstein, Almut G; Univ Arizona, Coll Pharm, Dept Pharm Practice & Sci (AMER MEDICAL ASSOC, 2019-12-02)
      IMPORTANCE Calcium channel blockers, specifically dihydropyridine calcium channel blockers (DH CCBs, eg, amlodipine), may cause lower-extremity edema. Anecdotal reports suggest this may result in a prescribing cascade, where DH CCB-induced edema is treated with loop diuretics. OBJECTIVE To assess the magnitude and characteristics of the DH CCB prescribing cascade. DESIGN, SETTING, AND PARTICIPANTS This cohort study used a prescription sequence symmetry analysis to assess loop diuretic initiation before and after the initiation of DH CCBs among patients aged 20 years or older without heart failure. Data from a private insurance claims database from 2005 to 2017 was analyzed. Use of loop diuretics associated with initiation of angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, and other commonly used medications was used as negative controls. Data were analyzed from March 2019 through October 2019. EXPOSURES Initiation of DH CCB or negative control medications. MAIN OUTCOMES AND MEASURES The temporality of loop diuretic initiation relative to DH CCB or negative control initiation. Secular trend-adjusted sequence ratios (aSRs) with 95% CIs were calculated using data from 360 days before and after initiation of DH CCBs. RESULTS Among 1 206 093 DH CCB initiators, 55 818 patients (4.6%) (33 100 [59.3%] aged <65 years; 32 916 [59.0%] women) had a new loop diuretic prescription 360 days before or after DH CCB initiation, resulting in an aSR of 1.87 (95% CI, 1.84-1.90). An estimated 1.44% of DH CCB initiators experienced the prescribing cascade. The aSR was disproportionately higher among DH CCB initiators who were prescribed high doses (aSR, 2.20; 95% CI, 2.13-2.27), initiated amlodipine (aSR, 1.89; 95% CI, 1.86-1.93), were men (aSR, 1.96; 95% CI, 1.91-2.01), and used fewer antihypertensive classes (aSR, 2.55; 95% CI, 2.47-2.64). The evaluation of ACE inhibitors or ARBs as negative controls suggested hypertension progression may have tempered the incidence of the prescribing cascade (aSR for ACE inhibitors and ARBs, 1.27; 95% CI, 1.24-1.29). CONCLUSIONS AND RELEVANCE This study found an excessive use of loop diuretics following initiation of DH CCBs that cannot be completely explained by secular trends or hypertension progression. The prescribing cascade was more pronounced among those initially prescribed a high dose of DH CCBs.
    • Prevalence of Female Authors in Case Reports Published in the Medical Literature

      Hsiehchen, David; Hsieh, Antony; Espinoza, Magdalena; Univ Arizona, Banner Univ Med Ctr (AMER MEDICAL ASSOC, 2019-05-31)
      IMPORTANCE Under representation of female authors in research publications is prevalent, but it is unclear whether this is attributable to sex disparities in research conduct or authorship practices. Case reports are a poorly understood component of the biomedical corpus, and the production of anecdotal observations is not confounded by factors associated with disparities in female representation in research publications. Whether female authorship disparities exist in nonresearch publications of clinical information is unknown. OBJECTIVES To examine the authorship of case reports and elucidate factors associated with sex disparity. DESIGN AND SETTING Cross-sectional study of all case reports published by US authors in 2014 and 2015 indexed in PubMed performed from July 2015 to July 2018. MAIN OUTCOMES AND MEASURES The primary outcome measure was the proportion of female first authors. The secondary outcome measures were the proportion of female last authors and female authorship representation among different clinical specialties. RESULTS Bibliometric data was abstracted from 20 427 case reports published across 2538 journals. A total of 7252 (36%) and 4825 (25%) case reports had a female first and last author, respectively. In comparison, 44% and 34% of US trainees and physicians, respectively, were female in 2015. Among adult case reports, female authorship was more prevalent in academic environments compared with community settings (34.0% vs 28.2% for female first authors and 23.4% vs 19.7% for female last authors). Across states, the proportions of female first authors and last authors were universally less than the proportions of female trainees and active female physicians, respectively. Female first authorship was associated with larger author teams (odds ratio [OR], 1.02; 95% CI, 1.01-1.03), an academic affiliation (OR, 1.16; 95% CI, 1.06-1.27), and a female last author (OR, 1.58; 95% CI, 1.47-1.70). Relative to general internal medicine, specialties dominated by male clinicians were less frequently associated with female first authors. Several exceptions displaying a relatively equivalent tendency for male and female first authorship included oncology (OR, 0.97; 95% CI, 0.81-1.16), ophthalmology (OR, 0.87; 95% CI, 0.72-1.05), and radiation oncology (OR, 0.94 95% CI, 0.56-1.56). CONCLUSIONS AND RELEVANCE The underrepresentation of women among first and last authors in publications of case reports underscores the pervasiveness of sex disparities in medicine. Collaboration and female mentors may be critical instruments in upsetting longstanding practices associated with sex bias. Not all clinical specialties were associated with lower-than-expected female authorship, and further exploration of specialty-specific norms in publication and mentorship may elucidate specific barriers to female authorship.
    • Trends and Costs Associated With Suboptimal Physical Activity Among US Women With Cardiovascular Disease

      Okunrintemi, Victor; Benson, Eve-Marie A; Tibuakuu, Martin; Zhao, Di; Ogunmoroti, Oluseye; Valero-Elizondo, Javier; Gulati, Martha; Nasir, Khurram; Michos, Erin D; Univ Arizona, Coll Med, Div Cardiol (AMER MEDICAL ASSOC, 2019-04-12)
      IMPORTANCE Cardiovascular disease (CVD) is the leading cause of death and disability among women. Achievement of recommended physical activity (PA) levels is an essential component of CVD management. OBJECTIVE To describe trends, sociodemographic factors, and health care expenditures associated with suboptimal PA among a nationally representative sample of US women with CVD. DESIGN, SETTING, AND PARTICIPANTS This cross-sectional study used serial data from the Medical Expenditure Panel Survey from 2006 through 2015. The analyses were conducted in August 2018. Women who had self-reported and/or International Classification of Diseases, Ninth Revision, diagnosis of CVD were included. MAIN OUTCOMES AND MEASURES Recommended PA was defined as 30 minutes or more of moderate-to vigorous-intensity exercise, 5 or more days per week. Weighted logistic regression was used to examine the associations of various sociodemographic factors with suboptimal PA, adjusted for comorbidities. A 2-part econometric model was used to assess health care expenditures. RESULTS A total of 18 027 women were included in this study. The results were weighted to provide estimates for approximately 19.5 million adult women in the United States with CVD (mean [SD] age, 60.4 [16.9] years). More than half of the women with CVD reported suboptimal PA, a trend that increased during the 10-year period, with 58.2%(95% CI, 55.9%-60.5%) of participants reporting suboptimal PA in 2006-2007 vs 61.9%(95% CI, 59.7%-64.2%) in 2014-2015 (P = .004). The proportion of women with suboptimal PA differed by sociodemographic factors. In adjusted models, compared with non-Hispanic white women, African American women (odds ratio, 1.22; 95% CI, 1.08-1.38) and Hispanic women (odds ratio, 1.33; 95% CI, 1.13-1.58) were more likely to have suboptimal PA. Women from low-or very low-income strata (compared with high-income strata), enrolled in public insurance (compared with private insurance), and with less than high school education (compared with at least some college education) were more likely to have suboptimal PA. Health care costs among women with CVD with suboptimal PA were higher compared with those among women who met the recommended PA, and this increased through time, from a mean total health care expenditure of $12 724 (95% CI, $11 627-$13 821) in 2006-2007 to $14 820 (95% CI, $13 521-$16 119) in 2014-2015. CONCLUSIONS AND RELEVANCE The proportion of women with CVD not meeting recommended PA is high and increasing, particularly among certain racial/ethnic and socioeconomic groups, and is associated with significant health care costs. More must be done to improve PA for secondary prevention and reduction of expenditures among women with CVD.