The effect of mailed outreach on FIT completion among patients aged 45-50 in a safety net healthcare system.

PURPOSE: Colorectal cancer screening is recommended starting at age 45, but there has been little research on strategies to promote screening in patients younger than 50. METHODS: An outreach program quasi-randomly assigned patients aged 45-50 without recent fecal immunochemical test (FIT), colonoscopy or contraindications to screening to two intervention arms: electronic outreach with email and text (electronic outreach only) versus electronic outreach plus mailed outreach with FIT, an instructional letter and a prepaid return envelope (mailed + electronic outreach). In response to known disparities in screening uptake, all Black patients were assigned to receive mailed + electronic outreach. RESULTS: Among patients quasi-randomly assigned to an intervention (non-Black patients), the 180-day FIT completion rate was 18.8% in the electronic outreach only group (n = 1,318) and 25.0% in the mailed + electronic outreach group (n = 1,364) (difference 6.2% [95% CI 3.0, 9.4]). FIT completion was 16.6% among Black patients (n = 469), 8.4% (95% CI 4.1, 12.6) lower than among non-Black patients also assigned to mailed + electronic outreach. CONCLUSION: Among patients aged 45-50, mailed + electronic outreach had a greater effect on FIT completion than electronic outreach alone. Crossover between intervention groups likely lead to an underestimation of the effect of mailed outreach.

Delivering Genetic Testing for Patients with Prostate Cancer: Moving Beyond Provider Knowledge as a Barrier to Care.

INTRODUCTION: The 2018 National Comprehensive Cancer Network guidelines for prostate cancer genetic testing expanded access to genetic services. Few studies have examined how this change has affected provider practice outside of large cancer centers. METHODS: We conducted a qualitative study of multi-disciplinary health care providers treating patients with prostate cancer at a safety-net hospital. Participants completed an interview that addressed knowledge, practices, and contextual factors related to providing genetic services to patients with prostate cancer. A thematic analysis using both inductive and deductive coding was undertaken. RESULTS: Seventeen providers completed interviews. Challenges in identifying eligible patients for genetic testing stemmed from a lack of a) systems that facilitate routine patient identification, and b) readily available family history data for eligibility determination. Providers identified non-medical patient characteristics that influenced their referral process, including health literacy, language, cultural beliefs, patient distress, and cost. Providers who see patients at different times along the cancer care continuum viewed benefits of testing differently. CONCLUSION: The use of digital technologies that systematically identify those eligible for genetic testing referrals may mitigate some but not all challenges identified in this study. Further research should determine how individual provider perceptions influence referral practices and patient access to genetics both within and across cancer specialties.

Warfarin Monitoring in Safety-Net Health Systems: Analysis by Race/Ethnicity and Language Preference.

BACKGROUND: Racial/ethnic disparities in anticoagulation management are well established. Differences in warfarin monitoring can contribute to these disparities and should be measured. OBJECTIVE: We assessed for differences in international normalized ratio (INR) monitoring by race/ethnicity and language preference across safety-net care systems serving predominantly low-income, ethnically diverse populations. DESIGN: Cross-sectional analysis of process and safety data shared from the Safety Promotion Action Research and Knowledge Network (SPARK-Net) initiative, a consortium of five California safety-net hospital systems. PARTICIPANTS: Eligible patients were at least 18 years old, received warfarin for at least 56 days during the measurement period from July 2015 to June 2017, and had INR testing in an ambulatory care setting at a participating healthcare system. MAIN MEASURES: We conducted a scaled Poisson regression for adjusted rate ratio of having at least one INR checked per 56-day time period for which a patient had a warfarin prescription. Adjusting for age, sex, healthcare system, and insurance status/type, we assessed for racial/ethnic and language disparities in INR monitoring. KEY RESULTS: Of 8129 patients, 3615 (44%) were female; 1470 (18%), Black/African American; 3354 (41%), Hispanic/Latinx; 1210 (15%), Asian; 1643 (20%), White; and 452 (6%), other. Three thousand five hundred forty-nine (45%) were non-English preferring. We did not observe statistically significant disparities in the rate of appropriate INR monitoring by race/ethnicity or language; the primary source of variation was by healthcare network. Older age, female gender, and uninsured patients had a slightly higher rate of appropriate INR monitoring, but differences were not clinically significant. CONCLUSIONS: We did not find a race/ethnicity nor language disparity in INR monitoring; safety-net site was the main source of variation.

Advocacy in Action: a Comprehensive Student-Led Proactive Outreach to Patients at Highest Risk.

INTRODUCTION: COVID-19 disrupted access to critical healthcare and resources for many, especially affecting patients at safety-net hospitals who rely on regular care for multiple complex conditions. Students realized they could support patients from the sidelines by helping navigate abrupt healthcare changes and proactively addressing needs at home. AIM: To comprehensively identify and meet the clinical and social needs of Atlanta, Georgia's patients at highest risk, left without their usual access to healthcare, through proactive telephonic outreach. SETTING AND PATIENTS: Medical and Physician's Assistant students from Emory and Morehouse Schools of Medicine partnered with Grady Health System, Atlanta's safety-net hospital. Artificial intelligence prioritized over 15,000 patients by risk of morbidity and mortality from COVID-19. PROGRAM DESCRIPTION: In this novel program, students performed telephonic outreach to thousands of patients at highest risk of poor outcomes from COVID-19. Students used a custom REDCap form that served as both a call script and data collection tool. It provided step-by-step guidance to (1) screen for COVID-19 and educate on prevention; (2) help patients navigate health system changes to fill gaps in care; and (3) identify and address social needs. Based on patients' responses, the form prompted tailored reminders for next steps and connections to medical and social resources. PROGRAM EVALUATION: In the program's first 16 months, students made 7,988 calls, of which 3,354 were answered. Over half (53%) of patients had at least one need requiring action: 48% health and 16% social. DISCUSSION: This proactive, novel initiative identified substantial clinical and social need among patients at highest risk for poor outcomes and filled a pressing health system gap exacerbated by COVID-19. Simultaneously, interprofessional students gained applied exposure to health systems sciences. This program can serve as a model for rapid, cost-effective, high-yield outreach to promote patient health at home both during and beyond the pandemic.

Universal social needs assessment in gynecologic oncology: An important step toward more informed and targeted care in the public safety net.

BACKGROUND: Social needs are actionable mediators of social determinants of health. Along with distress, they affect quality of life and survival in patients with cancer. The objectives of this study were to identify the most common social needs and distress in a largely immigrant gynecologic oncology patient population at a public safety-net hospital and to evaluate for specific needs associated with distress and poor outcomes. METHODS: This was a prospective, survey-based cohort study of patients who participated in a performance-improvement initiative offering social needs assessment and distress screening. Patients provided sociodemographic information and completed validated surveys adapted from the Health Leads Social Needs Screening Toolkit, the National Comprehensive Cancer Network Distress Thermometer, and the Emotion Thermometers Tool. Associations between social needs, distress, and treatment outcomes were analyzed. RESULTS: In total, 135 women were included. Of these, 65.2% had at least 1 unmet social need, and 36.3% screened positive for distress. Help reading hospital materials (30.4%) was the most frequently reported need. Social isolation (odds ratio [OR], 3.65; 95% CI, 1.35-9.9; P = .01) and lack of safety at home (OR, 4.90; 95% CI, 2.23-10.62; P = .0001) were associated with distress. Perceived lack of finances for medical care (OR, 5.69; 95% CI, 1.12-28.9; P = .036) and lack of transportation (OR, 20.5; 95% CI, 2.69-156.7; P = .004) were associated with nonadherence-related treatment interruption, whereas positive distress scores were associated with interruption because of comorbidities or treatment-related toxicities (OR, 20.5; 95% CI, 1.5-268.6; P = .02). CONCLUSIONS: Systematically identifying social needs and developing interventions aimed at mitigating them may lead to more actionable health care disparities research and affect treatment outcomes. LAY SUMMARY: Social needs are individual-level social conditions that drive health disparities. In this survey-based study, the objective was to identify common social needs and how these relate to distress and poor health outcomes in a largely immigrant and underserved gynecologic oncology patient population. The authors found that greater than one-third of patients screened positive for distress, nearly two-thirds had at least 1 unmet social need, and these factors were associated with emergency room visits, hospital admissions, and treatment interruptions. These findings suggest that screening for universal social needs allows providers to identify unrecognized needs and implement interventions to mitigate distress and improve health outcomes.

The PREPARE for Your Care program increases advance care planning engagement among diverse older adults with cancer.

BACKGROUND: Advance care planning (ACP) is low among older adults with cancer. In a secondary analysis of randomized trial data, the authors compared the efficacy of the PREPARE for Your Care (PREPARE) website plus an easy-to-read advance directive (AD) with an AD only among older adults with and without cancer. METHODS: Safety net, primary care patients in San Francisco were included if they were 55 years old or older, were English- or Spanish-speaking, and had 2 or more chronic conditions. The authors determined cancer diagnoses by using International Classification of Diseases, Ninth Revision/Tenth Revision codes. The primary outcome was new ACP documentation in the medical record at 15 months; the secondary outcomes were self-reported ACP engagement, ease of use, satisfaction, and depression/anxiety. The authors used mixed effects logistic and linear regression adjusted for prior ACP, health literacy, and clinician, including a cancer interaction term. RESULTS: Of 986 participants, 220 (22%) had cancer. The mean age was 63 years (SD, 6 years), 61% were women, 81% were of a minority race/ethnicity, 45% were Spanish-speaking, 39% had limited health literacy, and 27% had prior ACP. New ACP documentation was higher in the PREPARE arm versus the AD-only arm among participants with cancer (62% vs 43%; P = .01) and without cancer (38% vs 28%; P = .01), as was ACP engagement in both arms (P < .001), with no interactions by cancer. Ease of use and satisfaction were high, and depression/anxiety was low, with no differences by study arm or by cancer/no cancer. CONCLUSIONS: PREPARE plus an easy-to-read AD increased ACP documentation and engagement among diverse older adults with cancer more than an AD alone, with no increase in depression or anxiety between study arms or by cancer. PREPARE may help to decrease ACP disparities among patients with cancer. LAY SUMMARY: Advance care planning (ACP) is the process of sharing values, goals, and preferences for medical care, but engagement in ACP is low among older adults with cancer. Among 986 English- and Spanish-speaking older adults from a safety net hospital, an interactive, multimedia, web-based ACP program (PREPARE for Your Care at https://prepareforyourcare.org/) plus an easy-to-read advance directive increased ACP documentation and engagement more than an advance directive alone. There were no differences in this increase in ACP between older adults with cancer and older adults without cancer. Also, engaging in ACP did not result in increased depression or anxiety.

Man vs. Machine: Comparing Physician vs. Electronic Health Record-Based Model Predictions for 30-Day Hospital Readmissions.

BACKGROUND: Electronic health record (EHR)-based readmission risk prediction models can be automated in real-time but have modest discrimination and may be missing important readmission risk factors. Clinician predictions of readmissions may incorporate information unavailable in the EHR, but the comparative usefulness is unknown. We sought to compare clinicians versus a validated EHR-based prediction model in predicting 30-day hospital readmissions. METHODS: We conducted a prospective survey of internal medicine clinicians in an urban safety-net hospital. Clinicians prospectively predicted patients' 30-day readmission risk on 5-point Likert scales, subsequently dichotomized into low- vs. high-risk. We compared human with machine predictions using discrimination, net reclassification, and diagnostic test characteristics. Observed readmissions were ascertained from a regional hospitalization database. We also developed and assessed a "human-plus-machine" logistic regression model incorporating both human and machine predictions. RESULTS: We included 1183 hospitalizations from 106 clinicians, with a readmission rate of 20.8%. Both clinicians and the EHR model had similar discrimination (C-statistic 0.66 vs. 0.66, p = 0.91). Clinicians had higher specificity (79.0% vs. 48.9%, p < 0.001) but lower sensitivity (43.9 vs. 75.2%, p < 0.001) than EHR model predictions. Compared with machine, human was better at reclassifying non-readmissions (non-event NRI + 30.1%) but worse at reclassifying readmissions (event NRI - 31.3%). A human-plus-machine approach best optimized discrimination (C-statistic 0.70, 95% CI 0.67-0.74), sensitivity (65.5%), and specificity (66.7%). CONCLUSION: Clinicians had similar discrimination but higher specificity and lower sensitivity than EHR model predictions. Human-plus-machine was better than either alone. Readmission risk prediction strategies should incorporate clinician assessments to optimize the accuracy of readmission predictions.

Patient Factors Limit Colon Cancer Survival at Safety-Net Hospitals: A National Analysis.

BACKGROUND: Safety-net hospitals serve a vital role in society by providing care for vulnerable populations. Existing data regarding oncologic outcomes of patients with colon cancer treated at safety-net hospitals are limited and variable. The objective of this study was to delineate disparities in treatment and outcomes for patients with colon cancer treated at safety-net hospitals. METHODS: This retrospective cohort study identified 802,304 adult patients with colon adenocarcinoma from the National Cancer Database between 2004-2016. Patients were stratified according to safety-net burden of the treating hospital as previously described. Patient, tumor, facility, and treatment characteristics were compared between groups as were operative and short-term outcomes. Cox proportional hazards regression was utilized to compare overall survival between patients treated at high, medium, and low burden hospitals. RESULTS: Patients treated at safety-net hospitals were demographically distinct and presented with more advanced disease. They were also less likely to receive surgery, adjuvant chemotherapy, negative resection margins, adequate lymphadenectomy, or a minimally invasive operative approach. On multivariate analysis adjusting for patient and tumor characteristics, survival was inferior for patients at safety-net hospitals, even for those with stage 0 (in situ) disease. CONCLUSION: This analysis revealed inferior survival for patients with colon cancer treated at safety-net hospitals, including those without invasive cancer. These findings suggest that unmeasured population differences may confound analyses and affect survival more than provider or treatment disparities.

Defining safety net hospitals in the health services research literature: a systematic review and critical appraisal.

BACKGROUND: The aim of this study was to identify the range of ways that safety net hospitals (SNHs) have been empirically operationalized in the literature and determine the extent to which patterns could be identified in the use of empirical definitions of SNHs. METHODS: We conducted a PRISMA guided systematic review of studies published between 2009 and 2018 and analyzed 22 articles that met the inclusion criteria of hospital-level analyses with a clear SNH definition. RESULTS: Eleven unique SNH definitions were identified, and there were no obvious patterns in the use of a definition category (Medicaid caseload, DSH payment status, uncompensated care, facility characteristics, patient care mix) by the journal type where the article appeared, dataset used, or the year of publication. CONCLUSIONS: Overall, there is broad variability in the conceptualization of, and variables used to define, SNHs. Our work advances the field toward the development of standards in measuring, operationalizing, and conceptualizing SNHs across research and policy questions.

A Technical Assistance Curriculum for Expanding Sustainable School-Based Oral Health Programs in the Carolinas' Dental Safety Net.

PURPOSE: School-based oral health programs (SBOHPs) provide opportunities to address oral health inequities by providing convenient access points for care. No published guidelines on SBOHP implementation existed. Our work describes how philanthropic, public, and academic organizations partnered to support dental safety net providers with designing comprehensive SBOHPs in North and South Carolina. DESCRIPTION: A multi-sector leadership team was established to manage a new SBOHP philanthropic-funded grant program organized into two phases, Readiness and Implementation, with the former a 6-month planning period in preparation of the latter. Readiness included technical assistance (TA) delivered through coaching and 15 online learning modules organized in four domains: operations, finance, enabling services, and impact. Organizations could apply for implementation grants after successful TA completion. Process evaluation was used including a Readiness Stoplight Report for tracking progression. ASSESSMENT: Ten Readiness grantees completed the TA. A variety of models resulted, including mobile, portable and fixed clinics. Descriptive analysis was conducted on the readiness stoplight reports. Components of the operation and finance domains required were the most time-intensive, specifically the development of policy manuals, production goals, and financial performance tracking. CONCLUSION: The program's structure resulted in (a) a two-state learning community, (b) SBOHP practice and policy alignment, and (c) coordinated program distribution. TA improvements are planned to account for COVID-19 threats, including school closures, space limitations, and transmission fears. Telehealth, non-aerosolizing procedures, and improved scheduling and communication can address concerns. Organizations considering SBOHPs should explore similar recommendations to navigate adverse circumstances.

Efficacy of Percutaneous Tibial Nerve Stimulation for Overactive Bladder in Women and Men at a Safety Net Hospital.

PURPOSE: We investigated efficacy and compliance related to percutaneous tibial nerve stimulation in patients treated for overactive bladder at a large, urban safety net hospital. MATERIALS AND METHODS: Consecutive patients who underwent percutaneous tibial nerve stimulation at Grady Memorial Hospital from May 2015 through January 2019 were included in our cohort and records were reviewed retrospectively. Primary outcomes of interest included self-reported urinary symptoms and episodes of urinary incontinence. Our secondary outcome of interest was patient compliance, defined as completion of 12 or more treatment sessions. Descriptive analysis and paired t-tests were performed. RESULTS: Of the 50 patients with a mean ± SD age of 59 ± 12 years 80% were black, 52% were male, 34% were uninsured and 54% subscribed to government insurance. Prior treatment included behavioral modification in 100% of cases, anticholinergics in 86% and mirabegron in 4%. Patients completed a mean of 10.7 ± 2.7 of the 12 planned weekly percutaneous tibial nerve stimulation treatments. Of the patients 70% completed all 12 weekly treatments and 77% of those who completed 12 treatments continued to maintenance treatment. After percutaneous tibial nerve stimulation treatment average symptoms improved across all metrics, including mean daytime frequency (from 11.0 to 6.6 episodes per day or -24.5%), nighttime frequency (from 4.8 to 2.5 episodes per night or -47.1%), urgency score (from 3.4 to 1.9 or -42.1%) and incontinence (from 1.6 to 0.4 episodes per day or -79.6%) (each p <0.001). A total of 43 patients (86%) reported symptom improvement. CONCLUSIONS: Percutaneous tibial nerve stimulation had favorable efficacy and compliance in a traditionally underserved patient population. This should be considered as a feasible modality to manage overactive bladder symptoms in patients in a similar demographic.

Association of Patient Social, Cognitive, and Functional Risk Factors with Preventable Hospitalizations: Implications for Physician Value-Based Payment.

BACKGROUND: Ambulatory care-sensitive condition (ACSC) hospitalizations are used to evaluate physicians' performance in Medicare value-based payment programs. However, these measures may disadvantage physicians caring for vulnerable populations because they omit social, cognitive, and functional factors that may be important determinants of hospitalization. OBJECTIVE: To determine whether social, cognitive, and functional risk factors are associated with ACSC hospitalization rates and whether adjusting for them changes outpatient safety-net providers' performance. DESIGN: Using data from the Medicare Current Beneficiary Survey, we conducted patient-level multivariable regression to estimate the association (as incidence rate ratios (IRRs)) between patient-reported social, cognitive, and functional risk factors and ACSC hospitalizations. We compared outpatient safety-net and non-safety-net providers' performance after adjusting for clinical comorbidities alone and after additional adjustment for social, cognitive, and functional factors captured in survey data. SETTING: Safety-net and non-safety-net clinics. PARTICIPANTS: Community-dwelling Medicare beneficiaries contributing 38,616 person-years from 2006 to 2013. MEASUREMENTS: Acute and chronic ACSC hospitalizations. RESULTS: After adjusting for clinical comorbidities, Alzheimer's/dementia (IRR 1.30, 95% CI 1.02-1.65), difficulty with 3-6 activities of daily living (ADLs) (IRR 1.43, 95% CI 1.05-1.94), difficulty with 1-2 instrumental ADLs (IADLs, IRR 1.54, 95% CI 1.26-1.90), and 3-6 IADLs (IRR 1.90, 95% CI 1.49-2.43) were associated with acute ACSC hospitalization. Low income (IRR 1.28, 95% CI 1.03-1.58), lack of educational attainment (IRR 1.33, 95% CI 1.04-1.69), being unmarried (IRR 1.18, 95% CI 1.01-1.36), difficulty with 1-2 IADLs (IRR 1.30, 95% CI 1.05-1.60), and 3-6 IADLs (IRR 1.44, 95% CI 1.16-1.80) were associated with chronic ACSC hospitalization. Adding these factors to standard Medicare risk adjustment eliminated outpatient safety-net providers' performance gap (p < .05) on ACSC hospitalization rates relative to non-safety-net providers. CONCLUSIONS: Social, cognitive, and functional risk factors are independently associated with ACSC hospitalizations. Failure to account for them may penalize outpatient safety-net providers for factors that are beyond their control.

Screening initiation with FIT or colonoscopy: Post-hoc analysis of a pragmatic, randomized trial.

Screening with FIT or colonoscopy can reduce CRC mortality. In our pragmatic, randomized trial of screening outreach over three years, patients annually received mailed FITs or colonoscopy invitations. We examined screening initiation after each mailing and crossover from the invited to other modality. Eligible patients (50-64 years, ≥1 primary-care visit before randomization, and no history of CRC) received mailed FIT kits (n = 2400) or colonoscopy invitations (n = 2400) from March 2013 through July 2016. Among those invited for colonoscopy, we used multinomial logistic regression to identify factors associated with screening initiation with colonoscopy vs. FIT vs. no screening after the first mailing. Most patients were female (61.8%) and Hispanic (48.9%) or non-Hispanic black (24.0%). Among those invited for FIT, 56.6% (n = 1359) initiated with FIT, whereas 3.3% (n = 78) crossed over to colonoscopy; 151 (15.7%) and 61 (7.7%) initiated with FIT after second and third mailings. Among those invited for colonoscopy, 25.5% (n = 613) initiated with colonoscopy whereas 18.8% (n = 452) crossed over to FIT; 112 (8.4%) and 48 (4.2%) initiated with colonoscopy after second and third mailings. Three or more primary-care visits prior to randomization were associated with initiating with colonoscopy (OR 1.49, 95% CI 1.17-1.91) and crossing over to FIT (OR 1.63, 95% CI 1.19-2.23). Although nearly half of patients initiated screening after the first mailing, few non-responders in either outreach group initiated after a second or third mailing. More patients invited to colonoscopy crossed over to FIT than those assigned to FIT crossed over to colonoscopy.

Clinic Workload, the Quality of Staff Relationships and Diabetes Management in Community Health Centers Catering to Latino and Chinese Patients.

We examine whether workplace climate-quality of staff relationships (QSR) and manageable clinic workload (MCW) are related to better patient care experiences and diabetes care in community health centers (CHCs) catering to Latino and Chinese patients. Patient experience surveys of adult patients with type 2 diabetes and workplace climate surveys of clinicians and staff from CHCs were included in an analytic sample. Comparisons of means analyses examine patient and provider characteristics. The associations of QSR, MCW and the diabetes care management were examined using regression analyses. Diabetes care process were more consistently provided in CHCs with high quality staff relations and more manageable clinic workload, but HbA1c, LDL cholesterol, and blood pressure outcomes were no different between clinics with high vs. low QSR and MCW. Focusing efforts on improvements in practice climate may lead to more consistent provision of important processes of diabetes care for these patients.

Neighborhood Socioeconomic Status and Primary Health Care: Usual Points of Access and Temporal Trends in a Major US Urban Area.

Neighborhood socioeconomic status (SES), an overall marker of neighborhood conditions, may determine residents' access to health care, independently of their own individual characteristics. It remains unclear, however, how the distinct settings where individuals seek care vary by neighborhood SES, particularly in US urban areas. With existing literature being relatively old, revealing how these associations might have changed in recent years is also timely in this US health care reform era. Using data on the Philadelphia region from 2002 to 2012, we performed multilevel analysis to examine the associations of neighborhood SES (measured as census tract median household income) with access to usual sources of primary care (physician offices, community health centers, and hospital outpatient clinics). We found no evidence that residence in a low-income (versus high-income) neighborhood was associated with poorer overall access. However, low-income neighborhood residence was associated with less reliance on physician offices [-4.40 percentage points; 95 % confidence intervals (CI) -5.80, -3.00] and greater reliance on the safety net provided by health centers [2.08; 95 % CI 1.42, 2.75] and outpatient clinics [1.61; 95 % CI 0.97, 2.26]. These patterns largely persisted over the 10 years investigated. These findings suggest that safety-net providers have continued to play an important role in ensuring access to primary care in urban, low-income communities, further underscoring the importance of supporting a strong safety net to ensure equitable access to care regardless of place of residence.

Testing an Adapted Modified Delphi Method: Synthesizing Multiple Stakeholder Ratings of Health Care Service Effectiveness.

BACKGROUND: The Affordable Care Act incentivizes health systems for better meeting patient needs, but often guidance about patient preferences for particular health services is limited. All too often vulnerable patient populations are excluded from these decision-making settings. A community-based participatory approach harnesses the in-depth knowledge of those experiencing barriers to health care. METHOD: We made three modifications to the RAND-UCLA appropriateness method, a modified Delphi approach, involving patients, adding an advisory council group to characterize existing knowledge in this little studied area, and using effectiveness rather than "appropriateness" as the basis for rating. As a proof of concept, we tested this method by examining the broadly delivered but understudied nonmedical services that community health centers provide. RESULTS: This method created discrete, new knowledge about these services by defining 6 categories and 112 unique services and by prioritizing among these services based on effectiveness using a 9-point scale. Consistent with the appropriateness method, we found statistical convergence of ratings among the panelists. DISCUSSION: Challenges include time commitment and adherence to a clear definition of effectiveness of services. This diverse stakeholder engagement method efficiently addresses gaps in knowledge about the effectiveness of health care services to inform population health management.

Are primary care providers prepared to care for survivors of breast cancer in the safety net?

BACKGROUND: With the growing number of survivors of breast cancer outpacing the capacity of oncology providers, there is pressure to transition patients back to primary care. Primary care providers (PCPs) working in safety-net settings may have less experience treating survivors, and little is known about their knowledge and views on survivorship care. The current study was performed to determine the knowledge, attitudes, and confidence of PCPs in the safety net at delivering care to survivors of breast cancer. METHODS: A modified version of the National Cancer Institute's Survey of Physician Attitudes Regarding Care of Cancer Survivors was given to providers at 2 county hospitals and 5 associated clinics (59 providers). Focus groups were held to understand barriers to survivorship care. RESULTS: Although the majority of providers believed PCPs have the skills necessary to provide cancer-related follow-up, the vast majority were not comfortable providing these services themselves. Providers were adherent to American Society of Clinical Oncology recommendations for mammography (98%) and physical examination (87%); less than one-third were guideline-concordant for laboratory testing and only 6 providers (10%) met all recommendations. PCPs universally requested additional training on clinical guidelines and the provision of written survivorship care plans before transfer. Concerns voiced in qualitative sessions included unfamiliarity with the management of endocrine therapy and confusion regarding who would be responsible for certain aspects of care. CONCLUSIONS: Safety-net providers currently lack knowledge of and confidence in providing survivorship care to patients with breast cancer. Opportunities exist for additional training in evidence-based guidelines and improved coordination of care between PCPs and oncology specialists.

Sustaining "meaningful use" of health information technology in low-resource practices.

PURPOSE: The implementation of electronic health records (EHRs) has been extensively studied, but their maintenance once implemented has not. The Regional Extension Center (REC) program provides implementation assistance to priority practices-those with limited financial, technical, and organizational resources-but the assistance is time limited. Our objective was to identify potential barriers to maintenance of meaningful use of EHRs in priority primary care practices using a qualitative observational study for federally qualified health centers (FQHCs) and priority practices in Michigan. METHODS: We conducted cognitive task analysis (CTA) interviews and direct observations of health information technology implementation in FQHCs. In addition, we conducted semistructured interviews with implementation specialists serving priority practices to detect emergent themes relevant to maintenance. RESULTS: Maintaining EHR technology will require ongoing expert technical support indefinitely beyond implementation to address upgrades and security needs. Maintaining meaningful use for quality improvement will require ongoing support for leadership and change management. Priority practices not associated with larger systems lack access to the necessary technical expertise, financial resources, and leverage with vendors to continue alone. Rural priority practices are particularly challenged, because expertise is often not available locally. CONCLUSIONS: Priority practices, especially in rural areas, are at high risk for falling on the wrong side of a "digital divide" as payers and regulators enact increasing expectations for EHR use and information management. For those without affiliation to maintain the necessary expert staff, ongoing support will be needed for those practices to remain viable.

Barriers and Facilitators to Online Portal Use Among Patients and Caregivers in a Safety Net Health Care System: A Qualitative Study.

BACKGROUND: Patient portals have the potential to support self-management for chronic diseases and improve health outcomes. With the rapid rise in adoption of patient portals spurred by meaningful use incentives among safety net health systems (a health system or hospital providing a significant level of care to low-income, uninsured, and vulnerable populations), it is important to understand the readiness and willingness of patients and caregivers in safety net settings to access their personal health records online. OBJECTIVE: To explore patient and caregiver perspectives on online patient portal use before its implementation at San Francisco General Hospital, a safety net hospital. METHODS: We conducted 16 in-depth interviews with chronic disease patients and caregivers who expressed interest in using the Internet to manage their health. Discussions focused on health care experiences, technology use, and interest in using an online portal to manage health tasks. We used open coding to categorize all the barriers and facilitators to portal use, followed by a second round of coding that compared the categories to previously published findings. In secondary analyses, we also examined specific barriers among 2 subgroups: those with limited health literacy and caregivers. RESULTS: We interviewed 11 patients and 5 caregivers. Patients were predominantly male (82%, 9/11) and African American (45%, 5/11). All patients had been diagnosed with diabetes and the majority had limited health literacy (73%, 8/11). The majority of caregivers were female (80%, 4/5), African American (60%, 3/5), caregivers of individuals with diabetes (60%, 3/5), and had adequate health literacy (60%, 3/5). A total of 88% (14/16) of participants reported interest in using the portal after viewing a prototype. Major perceived barriers included security concerns, lack of technical skills/interest, and preference for in-person communication. Facilitators to portal use included convenience, health monitoring, and improvements in patient-provider communication. Participants with limited health literacy discussed more fundamental barriers to portal use, including challenges with reading and typing, personal experience with online security breaches/viruses, and distrust of potential security measures. Caregivers expressed high interest in portal use to support their roles in interpreting health information, advocating for quality care, and managing health behaviors and medical care. CONCLUSIONS: Despite concerns about security, difficulty understanding medical information, and satisfaction with current communication processes, respondents generally expressed enthusiasm about portal use. Our findings suggest a strong need for training and support to assist vulnerable patients with portal registration and use, particularly those with limited health literacy. Efforts to encourage portal use among vulnerable patients should directly address health literacy and security/privacy issues and support access for caregivers.