Ovarian cancer risk among older patients with stable adnexal masses.

BACKGROUND: Few studies have evaluated the risk of cancer among older patients with stable adnexal masses in community-based settings to determine the duration of observation time needed. OBJECTIVE: This study aimed to assess the ovarian cancer risk among older patients with stable adnexal masses on ultrasound. STUDY DESIGN: This was a retrospective cohort study of patients in a large community-based health system aged ≥50 years with an adnexal mass <10 cm on ultrasound between 2016 and 2020 who had at least 1 follow-up ultrasound performed ≥6 weeks after initial ultrasound. Masses were considered stable on follow-up examination if they did not exhibit an increase of >1 cm in the greatest dimension or a change in standardized reported ultrasound characteristics. Ovarian cancer risk was determined at increasing time intervals of stability after initial ultrasound. RESULTS: Among 4061 patients with stable masses, the average age was 61 years (range, 50-99), with an initial mass size of 3.8 cm (range, 0.2-9.9). With a median follow-up of 3.7 years, 11 cancers were detected, with an absolute risk of 0.27%. Ovarian cancer risk declined with longer duration of stability, from 0.73 (95% confidence interval, 0.30-1.17) per 1000 person-years at 6 to 12 weeks, 0.63 (95% confidence interval, 0.19-1.07) at 13 to 24 weeks, 0.44 (95% confidence interval, 0.01-0.87) at 25 to 52 weeks, and 0.00 (95% confidence interval, 0.00-0.00) at >52 weeks. Expressed as number needed to reimage, ongoing ultrasound imaging would be needed for 369 patients whose masses show stability at 6 to 12 weeks, 410 patients at 13 to 24 weeks, 583 patients at 25 to 52 weeks, and >1142 patients with stable masses at 53 to 104 weeks to detect 1 case of ovarian cancer. CONCLUSION: In a diverse community-based setting, among patients aged ≥50 years with an adnexal mass that was stable for at least 6 weeks after initial ultrasound, the risk of ovarian cancer was very low at 0.27%. Longer demonstrated duration of stability was associated with progressively lower risk, with no cancer cases observed after 52 weeks of stability. These findings suggest that the benefit of ultrasound monitoring of stable masses beyond 12 months is minimal and may be outweighed by potential risks of repeated imaging.

Association of Surgical Timing with Outcomes in Early Stage Lung Cancer.

BACKGROUND: Optimal time to surgery for lung cancer is not well established. We aimed to assess whether time to surgery correlates with outcomes. METHODS: We assessed patients 18-84 years old who were diagnosed with stage I/II lung cancer at our integrated healthcare system from 2009 to 2019. Time to surgery was defined to start with disease confirmation (imaging or biopsy) prior to the surgery scheduling date. Outcomes of unplanned return to care within 30 days of lung cancer surgery, all-cause mortality, and disease recurrence were compared based on time to surgery before and after 2, 4, and 12 weeks. RESULTS: Of 2861 included patients, 70% were over 65 years old and 61% were female. Time to surgery occurred in 1-2 weeks for 6%, 3-4 weeks for 31%, 5-12 weeks for 58%, and 13-26 weeks for 5% of patients. Patients with time to surgery > 4 (vs. ≤ 4) weeks had greater risk of both death (hazard ratio (HR) 1.18, 95% confidence interval (CI) 1.00-1.39) and recurrence (HR 1.33, 95% CI 1.10-1.62). Associations were not statistically significant when dichotomizing time to surgery at 2 or 12 weeks for death (2 week HR 1.23, 95% CI 0.93-1.64; 12 week HR 1.35, 95% CI 0.97-1.88) and recurrence (2 week HR 1.54, 95% CI 0.85-2.80; 12 week HR 2.28, 95% CI 0.80-6.46). CONCLUSIONS: Early stage lung cancer patients with time to surgery within 4 weeks experienced lower rates of recurrence. Optimal time to surgical resection may be shorter than previously reported.

SARS-CoV-2 Clinical Outcomes in Patients with Cancer in a Large Integrated Health Care System in Northern California.

The SARS-CoV-2 (COVID-19) pandemic continues to affect many lives globally. Patients with cancer undergoing potentially immunosuppressive therapies appear to be at particular risk for the disease and its complications. Here, we describe the experience of patients with cancer within Kaiser Permanente, a large, integrated health system in Northern California. Between February 25, 2020, and June 8, 2020, 4,627 patients were diagnosed with COVID-19, of whom 33 had active cancer treatment within 180 days and 214 had a history of cancer. Patients with active cancer treatment had a statistically higher risk of requiring noninvasive ventilation (odds ratio [OR], 2.57; confidence interval [CI], 1.10-6.01), and there was a nonsignificant trend toward higher risk of death (OR, 2.78; CI, 0.92-8.43). Those with a history of cancer had comparable outcomes to those without cancer. These data demonstrate an increased risk of complications from COVID-19 for patients with active cancer treatment.

Prior Immunosuppressive Therapy and Severe Illness Among Patients Diagnosed with SARS-CoV-2: a Community-Based Study.

BACKGROUND: An estimated 10 million people in the USA are immunocompromised, a risk factor for severe COVID-19. Data informing whether immune-mediated medications lead to more severe infection are sparse. OBJECTIVE: Determine whether outpatient immunosuppressive therapies that treat autoimmune inflammatory disease or prevent solid organ transplant rejection are associated with severe illness after diagnosis with SARS-CoV-2 DESIGN: Retrospective cohort study PARTICIPANTS: Adults with a positive PCR nasal swab for SARS-CoV-2 from February 25 to September 9, 2020, cared for within a large integrated health care organization MAIN MEASURES: Exposure was defined as an outpatient fill of prednisone, immunomodulator, small-molecule, or biologic therapy in the 105 days prior to a positive SARS-CoV-2 PCR test. The main outcome was either hospitalization, ICU admission, or death within 45 days after diagnosis of SARS-CoV-2. Multivariable logistic regression models were adjusted for age, race, gender, body mass index, comorbidities, and autoimmune disease. KEY RESULTS: A total of 39,686 adults had a positive PCR test. In the primary analysis, prior prednisone use was associated with severe illness after diagnosis with SARS-CoV-2 (odds ratio (OR) 1.31; 95% confidence interval (CI) 1.08-1.60); however, immunomodulator (OR 0.88; 95% CI 0.57-1.34) and biologic/small-molecule therapy (OR 1.26; 95% CI 0.79-2.00) were not. Secondary analyses showed variable risk among therapies: Janus-kinase inhibitors had an increased odds of severe illness (OR 3.35; 95% CI 1.16-9.67), thiopurines/conventionaldisease-modifying antirheumatic drugs had a reduced odds (OR 0.53; 95% CI 0.32-0.88), and tumor necrosis factor inhibitors were not associated (OR 0.45; 95% CI 0.18-1.08). CONCLUSIONS AND RELEVANCE: Outpatient use of prednisone is associated with severe illness after diagnosis of SARS-CoV-2. Immunomodulator and biologic/small-molecule therapy were not associated, but different risk subgroups were identified. Our findings can inform risk-benefit discussions in the clinic and risk-based recommendations for patients on these therapies.

A Randomized Encouragement Trial to Increase Mail Order Pharmacy Use and Medication Adherence in Patients with Diabetes.

BACKGROUND: Mail order pharmacy (MOP) use has been linked to improved medication adherence and health outcomes among patients with diabetes. However, no large-scale intervention studies have assessed the effect of encouraging MOP use on medication adherence. OBJECTIVE: To assess an intervention to encourage MOP services to increase its use and medication adherence. DESIGN: Randomized encouragement trial. PATIENTS: 63,012 diabetes patients from three health care systems: Kaiser Permanente Northern California (KPNC), Kaiser Permanente Hawaii (KPHI), and Harvard Pilgrim Health Care (HPHC) who were poorly adherent to at least one class of cardiometabolic medications and had not used MOP in the prior 12 months. INTERVENTION: Patients were randomized to receive either usual care (control arm) or outreach encouraging MOP use consisting of a mailed letter, secure email message, and automated telephone call outlining the potential benefits of MOP use (intervention arm). HPHC intervention patients received the letter only. MEASUREMENTS: We compared the percentages of patients that began using MOP and that became adherent to cardiometabolic medication classes during a 12-month follow-up period. We also conducted a race/ethnicity-stratified analysis. RESULTS: During follow-up, 10.6% of intervention patients began using MOP vs. 9.3% of controls (p < 0.01); the percent of cardiometabolic medication delivered via mail was 42.1% vs. 39.8% (p < 0.01). Metformin adherence improved in the intervention arm relative to control at the two KP sites (52% vs. 49%, p < 0.01). Stratified analyses suggested a significant positive effect of the intervention in White (RR: 1.12, 95% CI: 1.03, 1.22) and Asian (RR: 1.30, 95% CI: 1.17, 1.45) patients. CONCLUSION: This pragmatic trial showed that simple outreach to encourage MOP modestly increased its use and improved adherence measured by refills to a key class of diabetes medications in some settings. Given its minimal cost, clinicians and health systems should consider outreach interventions to actively promote MOP use among diabetes patients. TRIAL REGISTRATION: ClinicalTrials.gov registration number: NCT02621476.

Exploiting nonsystematic covariate monitoring to broaden the scope of evidence about the causal effects of adaptive treatment strategies.

In studies based on electronic health records (EHR), the frequency of covariate monitoring can vary by covariate type, across patients, and over time, which can limit the generalizability of inferences about the effects of adaptive treatment strategies. In addition, monitoring is a health intervention in itself with costs and benefits, and stakeholders may be interested in the effect of monitoring when adopting adaptive treatment strategies. This paper demonstrates how to exploit nonsystematic covariate monitoring in EHR-based studies to both improve the generalizability of causal inferences and to evaluate the health impact of monitoring when evaluating adaptive treatment strategies. Using a real world, EHR-based, comparative effectiveness research (CER) study of patients with type II diabetes mellitus, we illustrate how the evaluation of joint dynamic treatment and static monitoring interventions can improve CER evidence and describe two alternate estimation approaches based on inverse probability weighting (IPW). First, we demonstrate the poor performance of the standard estimator of the effects of joint treatment-monitoring interventions, due to a large decrease in data support and concerns over finite-sample bias from near-violations of the positivity assumption (PA) for the monitoring process. Second, we detail an alternate IPW estimator using a no direct effect assumption. We demonstrate that this estimator can improve efficiency but at the potential cost of increase in bias from violations of the PA for the treatment process.

Predictors of Hyperkalemia and Hypokalemia in Individuals with Diabetes: a Classification and Regression Tree Analysis.

BACKGROUND: Both hyperkalemia and hypokalemia can lead to cardiac arrhythmias and are associated with increased mortality. Information on the predictors of potassium in individuals with diabetes in routine clinical practice is lacking. OBJECTIVE: To identify predictors of hyperkalemia and hypokalemia in adults with diabetes. DESIGN: Retrospective cohort study, with classification and regression tree (CART) analysis. PARTICIPANTS: 321,856 individuals with diabetes enrolled in four large integrated health care systems from 2012 to 2013. MAIN MEASURES: We used a single serum potassium result collected in 2012 or 2013. Hyperkalemia was defined as a serum potassium ≥ 5.5 mEq/L and hypokalemia as < 3.5 mEq/L. Predictors included demographic factors, laboratory measurements, comorbidities, medication use, and health care utilization. KEY RESULTS: There were 2556 hypokalemia events (0.8%) and 1517 hyperkalemia events (0.5%). In univariate analyses, we identified concordant predictors (associated with increased probability of both hyperkalemia and hypokalemia), discordant predictors, and predictors of only hyperkalemia or hypokalemia. In CART models, the hyperkalemia "tree" had 5 nodes and a c-statistic of 0.76. The nodes were defined by prior potassium results and eGFRs, and the 5 terminal "leaves" had hyperkalemia probabilities of 0.2 to 7.2%. The hypokalemia tree had 4 nodes and a c-statistic of 0.76. The hypokalemia tree included nodes defined by prior potassium results, and the 4 terminal leaves had hypokalemia probabilities of 0.3 to 17.6%. Individuals with a recent potassium between 4.0 and 5.0 mEq/L, eGFR ≥ 45 mL/min/1.73m2, and no hypokalemia in the previous year had a < 1% rate of either hypokalemia or hyperkalemia. CONCLUSIONS: The yield of routine serum potassium testing may be low in individuals with a recent serum potassium between 4.0 and 5.0 mEq/L, eGFR ≥ 45 mL/min/1.73m2, and no recent history of hypokalemia. We did not examine the effect of recent changes in clinical condition or medications on acute potassium changes.

Prompting Patients with Poorly Controlled Diabetes to Identify Visit Priorities Before Primary Care Visits: a Pragmatic Cluster Randomized Trial.

BACKGROUND: Most patients with diabetes do not meet all evidence-based goals of care, and many patients report poor communication and lack of involvement in decision-making during primary care visits. OBJECTIVE: To test the hypothesis that a "Pre-Visit Prioritization" secure email message could improve visit communication and glycemic control among patients with type 2 diabetes. DESIGN: We conducted a pragmatic, provider-randomized, multi-site clinical trial from March 2015 to October 2016 across 30 primary care practices within Kaiser Permanente Northern California (KPNC), a large integrated care delivery system. PARTICIPANTS: Eligible patients had at least 1 year of KPNC membership, type 2 diabetes with most recently measured hemoglobin A1c (HbA1c) > = 8.0%, and were registered users of the KPNC online patient portal. INTERVENTIONS: Patients in the intervention arm, upon booking an appointment, received a secure email through the KPNC online portal with a link to the EHR allowing them to submit their top one or two priorities prior to the visit. Control patients received usual care. MAIN MEASURES: Glycemic control; change in HbA1c 6 and 12 months after the initial visit; patient-reported outcomes related to patient-provider communication and patient care experiences. KEY RESULTS: During the study period, 1276 patients had at least one eligible visit. In post-visit surveys (n = 457), more intervention arm patients reported preparing questions for their visit (72% vs 63%, p = 0.048) and being given treatment choices to consider (81% vs 73%, p = 0.041). Patients in both arms had similar reductions in HbA1c over the 12-month study period (0.56% ± 1.45%), with no significant differences between arms. CONCLUSIONS: A "light touch" email-based pre-visit intervention resulted in improved measures of visit interaction but did not significantly improve glycemic control relative to usual care. Improving diabetes clinical outcomes through more effective primary care visits may require more intensive approaches to patient visit preparation. TRIAL REGISTRY: NCT02375932.

Targeted learning with daily EHR data.

Electronic health records (EHR) data provide a cost- and time-effective opportunity to conduct cohort studies of the effects of multiple time-point interventions in the diverse patient population found in real-world clinical settings. Because the computational cost of analyzing EHR data at daily (or more granular) scale can be quite high, a pragmatic approach has been to partition the follow-up into coarser intervals of pre-specified length (eg, quarterly or monthly intervals). The feasibility and practical impact of analyzing EHR data at a granular scale has not been previously evaluated. We start filling these gaps by leveraging large-scale EHR data from a diabetes study to develop a scalable targeted learning approach that allows analyses with small intervals. We then study the practical effects of selecting different coarsening intervals on inferences by reanalyzing data from the same large-scale pool of patients. Specifically, we map daily EHR data into four analytic datasets using 90-, 30-, 15-, and 5-day intervals. We apply a semiparametric and doubly robust estimation approach, the longitudinal Targeted Minimum Loss-Based Estimation (TMLE), to estimate the causal effects of four dynamic treatment rules with each dataset, and compare the resulting inferences. To overcome the computational challenges presented by the size of these data, we propose a novel TMLE implementation, the "long-format TMLE," and rely on the latest advances in scalable data-adaptive machine-learning software, xgboost and h2o, for estimation of the TMLE nuisance parameters.

Opportunities to encourage mail order pharmacy delivery service use for diabetes prescriptions: a qualitative study.

BACKGROUND: Medication non-adherence is a major contributor to poor outcomes in diabetes. Previous research has shown an association between use of mail order pharmacy delivery and better medication adherence, but little is known about the barriers and facilitators to mail order pharmacy use in diabetes patients. This qualitative study examined factors related to mail order pharmacy use versus traditional "brick and mortar" pharmacies to refill prescriptions. METHODS: We conducted four 90-min focus groups in 2016 among 28 diabetes patients in the Hawaii and Northern California regions of Kaiser Permanente, a large integrated health care delivery system. We queried participants on their preferred mode for refilling prescriptions and perceived barriers and facilitators of mail order pharmacy use. One researcher independently coded each focus group transcript, with two of these transcripts double-coded by a second researcher to promote reliability. We employed thematic analysis guided by the Capability, Opportunity, Motivation, and Behavior (COM-B) framework using NVivo 11 software. RESULTS: A total of 28 diabetes patients participated. Participants' average age was 64.1 years; 57% were female; and racial/ethnic backgrounds included Asian/Native Hawaiian/Pacific Islander (36%), Black/African-American (21%) Hispanic/Latino (7%), and non-Hispanic White (36%). Analysis uncovered 26 themes related to the decision to use mail order pharmacy, with each theme representing a barrier or facilitator mapped to the COM-B framework. Most themes (20/26) fell into the COM-B category of 'Opportunity.' Opportunity barriers to mail order pharmacy use included unpredictability of medication delivery date, concerns about mail security, and difficulty coordinating refill orders for multiple prescriptions. In contrast, facilitators included greater access and convenience (e.g., no need to wait in line or arrange transportation) compared to traditional pharmacies. Motivational facilitators to mail order pharmacy use included receiving a pharmacy benefit plan incentive of a free one-month supply of prescriptions. CONCLUSIONS: This study found that while patients with diabetes may benefit from mail order pharmacy use, they perceive numerous barriers to using the service. These findings will inform the design of interventions and quality improvement initiatives to increase mail order pharmacy use, which in turn may improve medication adherence and outcomes in diabetes patients, across health care systems.

Building Toward a Population-Based Approach to Diabetes Screening and Prevention for US Adults.

PURPOSE OF REVIEW: Evidence-based treatments for prediabetes can prevent and delay the development of type 2 diabetes in adults. In this review, we propose a framework for population-based diabetes prevention that links screening and prevention activities across key stakeholders. We also discuss gaps in current practice, while highlighting opportunities to improve diabetes screening and prevention efforts population-wide. RECENT FINDINGS: Awareness of diabetes risk is low, and many adults with prediabetes are not identified through existing screening efforts. Accumulating evidence and policies support expansion of the Diabetes Prevention Program (DPP) into clinical and community settings. However, the infrastructure to facilitate referrals and promote data exchange among patients, clinical settings, and community-based DPP programs is lacking. Development of evidence-driven, scalable processes for assessing diabetes risk, screening eligible adults, and delivering preventive treatments are needed to effectively improve the glycemic health of the US adult population.

Influence of a New Diabetes Diagnosis on the Health Behaviors of the Patient's Partner.

PURPOSE: When a person is given a diagnosis of diabetes, the changes in his or her health behaviors may influence the behaviors of his or her partner. The diabetes diagnosis may affect household members' perceptions of their own health risks, which could trigger behavioral change. The purpose of this study was to assess whether partners of persons with newly diagnosed diabetes changed their health behaviors compared with partners of persons without diabetes. METHODS: The study population consisted of Kaiser Permanente Northern California health plan members from 2007 to 2011. This cohort study assessed differences in change of 8 health behaviors. The study compared coresiding partners of persons with newly diagnosed diabetes before and after a diabetes diagnosis with a 5 to 1 matched sample of coresiding partners of persons without diabetes. RESULTS: A total of 180,910 couples were included in the analysis. After adjusting for baseline characteristics, partners of persons with newly diagnosed diabetes had significantly higher rates of participation in weight management-related health education classes (risk ratio [RR] = 1.50; 95% CI, 1.39-1.63); smoking cessation medication use (RR = 1.25; 95% CI, 1.05-1.50); glucose screening (RR = 1.07; 95% CI, 1.05-1.08); clinically meaningful weight loss (RR = 1.06; 95% CI, 1.02-1.11); lipid screening (RR = 1.05; 95% CI, 1.04-1.07); influenza vaccination (RR = 1.03; 95% CI, 1.02-1.04); and blood pressure screening (RR = 1.02; 95% CI, 1.02-1.03) compared with partners of persons without diabetes. CONCLUSIONS: There were small but significant differences in health-related behavioral changes among partners of persons with newly diagnosed diabetes compared with partners of persons without diabetes, even when no intervention occurred. This finding suggests a diabetes diagnosis within a family may be a teachable moment to improve health behaviors at the household level.

Population Health Management for Diabetes: Health Care System-Level Approaches for Improving Quality and Addressing Disparities.

PURPOSE OF REVIEW: Population care approaches for diabetes have the potential to improve the quality of care and decrease diabetes-related mortality and morbidity. Population care strategies are particularly relevant as accountable care organizations (ACOs), patient-centered medical homes (PCMH), and integrated delivery systems are increasingly focused on managing chronic disease care at the health system level. This review outlines the key elements of population care approaches for diabetes in the current health care environment. RECENT FINDINGS: Population care approaches proactively identify diabetes patients through disease registries and electronic health record data and utilize multidisciplinary care teams, personalized provider feedback, and decision support tools to target and care for patients at risk for poor outcomes. Existing evidence suggests that these strategies can improve care outcomes and potentially ameliorate existing race/ethnic disparities in health care. However, such strategies may be less effective for patients who are disengaged from the health care system. As population care for diabetes continues to evolve, future initiatives should consider ways to tailor population care to meet individual patient needs, while leveraging improvements in clinical information systems and care integration to optimally manage and prevent diabetes in the future.

Incidence of diabetes after a partner's diagnosis.

People socially connected with each other often share health risks, possibly due to shared environments and behaviors. In a cohort study, we examined whether incidence of diabetes was different for individuals with recently diagnosed partners compared to individuals similar on other characteristics but whose partners were never diagnosed with diabetes. We analyzed 2007-11 data from Kaiser Permanente Northern California (KPNC), an integrated health system with >3.5 million members. We estimated annual diabetes incidence controlling for demographic, socio-economic, behavioral, and health characteristics. Using propensity score matching and multivariate logistic regression, we compared odds of incident diabetes among co-residing partners ages 18-89years of people who had been diagnosed with diabetes during the previous year (in robustness checks up to the previous three years) and people who had never been diagnosed but were similar on observed characteristics. Partners of newly-diagnosed people had annual diabetes incidence of 16.4/1000, equivalent to10.8 times higher (95%CI: 9.2-12.6) than people whose spouses had never been diagnosed (1.5/1000). Odds remained higher three years after a spouses' diagnosis (45.4 vs. 11.7/1000). Adjusting for other characteristics, odds of diabetes for those with a partner diagnosed in the previous year were 8.7 times higher (CI: 7.4-10.2) than among those whose partner had never been diagnosed. Also, partners of persons with recently-diagnosed diabetes developed diabetes at much higher rates than the general KPNC and U.S. POPULATIONS: Individuals with a recently diagnosed partner could be considered a high-risk population for screening and prevention.

Medication Adherence Does Not Explain Black-White Differences in Cardiometabolic Risk Factor Control among Insured Patients with Diabetes.

BACKGROUND: Among patients with diabetes, racial differences in cardiometabolic risk factor control are common. The extent to which differences in medication adherence contribute to such disparities is not known. We examined whether medication adherence, controlling for treatment intensification, could explain differences in risk factor control between black and white patients with diabetes. METHODS: We identified three cohorts of black and white patients treated with oral medications and who had poor risk factor control at baseline (2009): those with glycated hemoglobin (HbA1c) >8 % (n = 37,873), low-density lipoprotein cholesterol (LDL-C) >100 mg/dl (n = 27,954), and systolic blood pressure (SBP) >130 mm Hg (n = 63,641). Subjects included insured adults with diabetes who were receiving care in one of nine U.S. integrated health systems comprising the SUrveillance, PREvention, and ManagEment of Diabetes Mellitus (SUPREME-DM) consortium. Baseline and follow-up risk factor control, sociodemographic, and clinical characteristics were obtained from electronic health records. Pharmacy-dispensing data were used to estimate medication adherence (i.e., medication refill adherence [MRA]) and treatment intensification (i.e., dose increase or addition of new medication class) between baseline and follow-up. County-level income and educational attainment were estimated via geocoding. Logistic regression models were used to test the association between race and follow-up risk factor control. Models were specified with and without medication adherence to evaluate its role as a mediator. RESULTS: We observed poorer medication adherence among black patients than white patients (p < 0.01): 50.6 % of blacks versus 39.7 % of whites were not highly adherent (i.e., MRA <80 %) to HbA1c oral medication(s); 58.4 % of blacks and 46.7 % of whites were not highly adherent to lipid medication(s); and 33.4 % of blacks and 23.7 % of whites were not highly adherent to BP medication(s). Across all cardiometabolic risk factors, blacks were significantly less likely to achieve control (p < 0.01): 41.5 % of blacks and 45.8 % of whites achieved HbA1c <8 %; 52.6 % of blacks and 60.8 % of whites achieved LDL-C <100; and 45.7 % of blacks and 53.6 % of whites achieved SBP <130. Adjusting for medication adherence/treatment intensification did not alter these patterns or model fit statistics. CONCLUSIONS: Medication adherence failed to explain observed racial differences in the achievement of HbA1c, LDL-C, and SBP control among insured patients with diabetes.

The Diabetes Telephone Study: Design and challenges of a pragmatic cluster randomized trial to improve diabetic peripheral neuropathy treatment.

BACKGROUND: Challenges to effective pharmacologic management of symptomatic diabetic peripheral neuropathy include the limited effectiveness of available medicines, frequent side effects, and the need for ongoing symptom assessment and treatment titration for maximal effectiveness. We present here the rationale and implementation challenges of the Diabetes Telephone Study, a randomized trial designed to improve medication treatment, titration, and quality of life among patients with symptomatic diabetic peripheral neuropathy. METHODS: We implemented a pragmatic cluster randomized controlled trial to test the effectiveness of an automated interactive voice response tool designed to provide physicians with real-time patient-reported data about responses to newly prescribed diabetic peripheral neuropathy medicines. A total of 1834 primary care physicians treating patients in the diabetes registry at Kaiser Permanente Northern California were randomized into the intervention or control arm. In September 2014, we began identification and recruitment of patients assigned to physicians in the intervention group who receive three brief interactive calls every 2 months after a medication is prescribed to alleviate diabetic peripheral neuropathy symptoms. These calls provide patients with the opportunity to report on symptoms, side effects, self-titration of medication dose and overall satisfaction with treatment. We plan to compare changes in self-reported quality of life between the intervention group and patients in the control group who receive three non-interactive automated educational phone calls. RESULTS: Successful implementation of this clinical trial required robust stakeholder engagement to help tailor the intervention and to address pragmatic concerns such as provider time constraints. As of 27 October 2015, we had screened 2078 patients, 1447 of whom were eligible for participation. We consented and enrolled 1206 or 83% of those eligible. Among those enrolled, 53% are women and the mean age is 67 (standard deviation = 12) years. The racial ethnic make-up is 56% White, 8% Asian, 13% Black or African American, and 19% Hispanic or Latino. CONCLUSION: Innovative strategies are needed to guide improvements in healthcare delivery for patients with symptomatic diabetic peripheral neuropathy. This trial aims to assess whether real-time collection and clinical feedback of patient treatment experiences can reduce patient symptom burden. Implementation of a clinical trial closely involving clinical care required researchers to partner with clinicians. If successful, this intervention provides a critical information feedback loop that would optimize diabetic peripheral neuropathy medication titration through widely available interactive voice response technology.

Health care system-level factors associated with performance on Medicare STAR adherence metrics in a large, integrated delivery system.

BACKGROUND: The Centers for Medicare and Medicaid Services provide significant incentives to health plans that score well on Medicare STAR metrics for cardiovascular disease risk factor medication adherence. Information on modifiable health system-level predictors of adherence can help clinicians and health plans develop strategies for improving Medicare STAR scores, and potentially improve cardiovascular disease outcomes. OBJECTIVE: To examine the association of Medicare STAR adherence metrics with system-level factors. RESEARCH DESIGN: A cross-sectional study. SUBJECTS: A total of 129,040 diabetes patients aged 65 years and above in 2010 from 3 Kaiser Permanente regions. MEASURES: Adherence to antihypertensive, antihyperlipidemic, and oral antihyperglycemic medications in 2010, defined by Medicare STAR as the proportion of days covered ≥ 80%. RESULTS: After controlling for individual-level factors, the strongest predictor of achieving STAR-defined medication adherence was a mean prescribed medication days' supply of > 90 days (RR=1.61 for antihypertensives, oral antihyperglycemics, and statins; all P < 0.001). Using mail order pharmacy to fill medications > 50% of the time was independently associated with better adherence with these medications (RR = 1.07, 1.06, 1.07; P < 0.001); mail order use had an increased positive association among black and Hispanic patients. Medication copayments ≤ $10 for 30 days' supply (RR = 1.02, 1.02, 1.02; P < 0.01) and annual individual out-of-pocket maximums ≤ $2000 (RR = 1.02, 1.01, 1.02; P < 0.01) were also significantly associated with higher adherence for all 3 therapeutic groupings. CONCLUSIONS: Greater medication days' supply and mail order pharmacy use, and lower copayments and out-of-pocket maximums, are associated with better Medicare STAR adherence. Initiatives to improve adherence should focus on modifiable health system-level barriers to obtaining evidence-based medications.

High-dimensional propensity score algorithm in comparative effectiveness research with time-varying interventions.

The high-dimensional propensity score (hdPS) algorithm was proposed for automation of confounding adjustment in problems involving large healthcare databases. It has been evaluated in comparative effectiveness research (CER) with point treatments to handle baseline confounding through matching or covariance adjustment on the hdPS. In observational studies with time-varying interventions, such hdPS approaches are often inadequate to handle time-dependent confounding and selection bias. Inverse probability weighting (IPW) estimation to fit marginal structural models can adequately handle these biases under the fundamental assumption of no unmeasured confounders. Upholding of this assumption relies on the selection of an adequate set of covariates for bias adjustment. We describe the application and performance of the hdPS algorithm to improve covariate selection in CER with time-varying interventions based on IPW estimation and explore stabilization of the resulting estimates using Super Learning. The evaluation is based on both the analysis of electronic health records data in a real-world CER study of adults with type 2 diabetes and a simulation study. This report (i) establishes the feasibility of IPW estimation with the hdPS algorithm based on large electronic health records databases, (ii) demonstrates little impact on inferences when supplementing the set of expert-selected covariates using the hdPS algorithm in a setting with extensive background knowledge, (iii) supports the application of the hdPS algorithm in discovery settings with little background knowledge or limited data availability, and (iv) motivates the application of Super Learning to stabilize effect estimates based on the hdPS algorithm.

Wellness Coaching for People With Prediabetes: A Randomized Encouragement Trial to Evaluate Outreach Methods at Kaiser Permanente, Northern California, 2013.

INTRODUCTION: Health coaching can improve lifestyle behaviors known to prevent or manage chronic conditions. Little is known about effective ways to encourage health and wellness coaching among people who might benefit. The purpose of this randomized encouragement trial was to assess the relative success of 3 outreach methods (secured email message, telephone message, and mailed letter) on the use of wellness coaching by people with prediabetes. METHODS: A total of 14,584 Kaiser Permanente Northern California (KPNC) patients with diagnosed prediabetes (fasting plasma glucose, 110-125mg/dL) were randomly assigned to be contacted via 1 of 4 intervention arms from January through May 2013. The uptake rate (making an appointment at the Wellness Coaching Center [WCC]) was assessed, and the association between uptake rate and patient characteristics was examined via multivariable logistic regression. RESULTS: The overall uptake rate across intervention arms was 1.9%. Secured email message had the highest uptake rate (3.0%), followed by letters and telephone messages (P < .05 for all pairwise comparisons). No participants in the usual-care arm (ie, no outreach) made an appointment with the WCC. For each year of increased age, the estimated odds of the uptake increased by 1.02 (odds ratio [OR] = 1.02; 95% CI, 1.01-1.04). Women were nearly twice as likely to make an appointment at the WCC as men (OR = 1.87; 95% CI, 1.40-2.51). CONCLUSION: Our results suggest that the WCC can recruit and encourage KPNC members with prediabetes to participate in the WCC. Future research should focus on increasing participation rates in health coaching among patients who may benefit.