Search

ObjectiveTo create and validate a simple and transferable machine learning model from electronic health record data to accurately predict clinical deterioration in patients with covid-19 across institutions, through use of a novel paradigm for model development and code sharing.DesignRetrospective cohort study.SettingOne US hospital during 2015-21 was used for model training and internal validation. External validation was conducted on patients admitted to hospital with covid-19 at 12 other US medical centers during 2020-21.Participants33 119 adults (≥18 years) admitted to hospital with respiratory distress or covid-19.Main outcome measuresAn ensemble of linear models was trained on the development cohort to predict a composite outcome of clinical deterioration within the first five days of hospital admission, defined as in-hospital mortality or any of three treatments indicating severe illness: mechanical ventilation, heated high flow nasal cannula, or intravenous vasopressors. The model was based on nine clinical and personal characteristic variables selected from 2686 variables available in the electronic health record. Internal and external validation performance was measured using the area under the receiver operating characteristic curve (AUROC) and the expected calibration error-the difference between predicted risk and actual risk. Potential bed day savings were estimated by calculating how many bed days hospitals could save per patient if low risk patients identified by the model were discharged early.Results9291 covid-19 related hospital admissions at 13 medical centers were used for model validation, of which 1510 (16.3%) were related to the primary outcome. When the model was applied to the internal validation cohort, it achieved an AUROC of 0.80 (95% confidence interval 0.77 to 0.84) and an expected calibration error of 0.01 (95% confidence interval 0.00 to 0.02). Performance was consistent when validated in the 12 external medical centers (AUROC range 0.77-0.84), across subgroups of sex, age, race, and ethnicity (AUROC range 0.78-0.84), and across quarters (AUROC range 0.73-0.83). Using the model to triage low risk patients could potentially save up to 7.8 bed days per patient resulting from early discharge.ConclusionA model to predict clinical deterioration was developed rapidly in response to the covid-19 pandemic at a single hospital, was applied externally without the sharing of data, and performed well across multiple medical centers, patient subgroups, and time periods, showing its potential as a tool for use in optimizing healthcare resources.

This national cohort survey of older adults assesses which health-related concerns they consider to be most salient in the 2024 election. [ABSTRACT FROM AUTHOR]

ImportancePatients with advanced-stage cancer are receiving increasingly aggressive medical care near death, despite growing concerns that this reflects poor-quality care.ObjectiveTo assess the association of aggressive end-of-life care with bereaved family members' perceptions of the quality of end-of-life care and patients' goal attainment.Design, setting, and participantsInterviews with 1146 family members of Medicare patients with advanced-stage lung or colorectal cancer in the Cancer Care Outcomes Research and Surveillance study (a multiregional, prospective, observational study) who died by the end of 2011 (median, 144.5 days after death; interquartile range, 85.0-551.0 days).ExposuresClaims-based quality measures of aggressive end-of-life care (ie, intensive care unit [ICU] admission or repeated hospitalizations or emergency department visits during the last month of life; chemotherapy ≤2 weeks of death; no hospice or ≤3 days of hospice services; and deaths occurring in the hospital).Main outcomes and measuresFamily member-reported quality rating of "excellent" for end-of-life care. Secondary outcomes included patients' goal attainment (ie, end-of-life care congruent with patients' wishes and location of death occurred in preferred place).ResultsOf 1146 patients with cancer (median age, 76.0 years [interquartile range, 65.0-87.0 years]; 55.8% male), bereaved family members reported excellent end-of-life care for 51.3%. Family members reported excellent end-of-life care more often for patients who received hospice care for longer than 3 days (58.8% [352/599]) than those who did not receive hospice care or received 3 or fewer days (43.1% [236/547]) (adjusted difference, 16.5 percentage points [95% CI, 10.7 to 22.4 percentage points]). In contrast, family members of patients admitted to an ICU within 30 days of death reported excellent end-of-life care less often (45.0% [68/151]) than those who were not admitted to an ICU within 30 days of death (52.3% [520/995]) (adjusted difference, -9.4 percentage points [95% CI, -18.2 to -0.6 percentage points]). Similarly, family members of patients who died in the hospital reported excellent end-of-life care less often (42.2% [194/460]) than those who did not die in the hospital (57.4% [394/686]) (adjusted difference, -17.0 percentage points [95% CI, -22.9 to -11.1 percentage points]). Family members of patients who did not receive hospice care or received 3 or fewer days were less likely to report that patients died in their preferred location (40.0% [152/380]) than those who received hospice care for longer than 3 days (72.8% [287/394]) (adjusted difference, -34.4 percentage points [95% CI, -41.7 to -27.0 percentage points]).Conclusions and relevanceAmong family members of older patients with fee-for service Medicare who died of lung or colorectal cancer, earlier hospice enrollment, avoidance of ICU admissions within 30 days of death, and death occurring outside the hospital were associated with perceptions of better end-of-life care. These findings are supportive of advance care planning consistent with the preferences of patients.

BackgroundAccurate information regarding race, ethnicity, and national origins is critical for identifying disparities in the cancer burden.ObjectivesTo examine the use of a Spanish surname list to improve the quality of race-related information obtained from rapid case ascertainment (RCA) and to estimate the accuracy of race-related information obtained from cancer registry records collected by routine reporting.SubjectsSelf-reported survey responses of 3954 participants from California enrolled in the Cancer Care Outcomes Research and Surveillance Consortium.MeasuresSensitivity, specificity, positive predictive value, and percent agreement. We used logistic regression to identify predictors of underreporting and overreporting of a race/ethnicity.ResultsUse of the Spanish surname list increased the sensitivity of RCA for Latino ethnicity from 37% to 83%. Sensitivity for cancer registry records collected by routine reporting was ≥95% for whites, blacks, and Asians, and specificity was high for all groups (86%-100%). However, patterns of misclassification by race/ethnicity were found that could lead to biased cancer statistics for specific race/ethnicities. Discordance between self-reported and registry-reported race/ethnicity was more likely for women, Latinos, and Asians.ConclusionsMethods to improve race and ethnicity data, such as using Spanish surnames in RCA and instituting data collection guidelines for hospitals, are needed to ensure minorities are accurately represented in clinical and epidemiological research.

BackgroundClinical trials are essential to establish the effectiveness of new cancer therapies, but less than 5% of adults with cancer enroll in trials. In addition to ineligibility or lack of available trials, barriers to enrollment may include limited patient awareness about the option of participation.MethodsWe surveyed a multiregional cohort of patients with lung or colorectal cancer (or their surrogates) three to six months after diagnosis. We assessed whether respondents reported learning that clinical trial participation might be an option, and, if so, with whom they discussed trials. We used logistic regression to assess the association of patient characteristics with discussing trial participation and enrolling in trials. All statistical tests were two-sided.ResultsOf 7887 respondents, 1114 (14.1%) reported discussing the possibility of clinical trial participation; most learned about trials from their physicians, and 287 patients (3.6% of all patients, 25.8% of trial discussants) enrolled. Among 2173 patients who received chemotherapy for advanced (stage III/IV lung or stage IV colorectal) cancer, 25.7% discussed trials, and 7.6% (29.5% of trial discussants) enrolled. Discussions were less frequent among older patients, African American or Asian vs white patients, and those with lower incomes and more comorbidity. Enrollment was higher among patients reporting shared vs physician-driven decisions (all P < .05).ConclusionsIn this population-based cohort, only 14% of patients discussed participation in clinical trials. Discussions were more frequent among advanced cancer patients but were still reported by a minority of patients. Strategies to expand access to trials and facilitate patient-provider communication about participation may accelerate development of better cancer therapeutics.

ObjectiveDisease-modifying antirheumatic drugs (DMARDs) are the standard of care for rheumatoid arthritis (RA); however, studies have found that many patients do not receive them. We examined predictors of starting and stopping DMARDs among a longitudinal cohort of patients with RA.MethodsStudy participants came from a cohort of RA patients recruited from a random sample of rheumatologists' practices in Northern California. We examined patterns and predictors of stopping and starting nonbiologic and biologic DMARDs during 1982-2009 based on annual questionnaires. Stopping was defined as stopping all DMARDs and starting was defined as transitioning from no DMARDs to any DMARDs across 2 consecutive years.ResultsThe analysis of starting DMARDs included 471 subjects with 1,974 pairs of years with no DMARD use in the first of 2 consecutive years. From this population, subjects started DMARD use by year 2 in 313 (15.9%) of the pairs. The analysis of stopping DMARDs included 1,026 subjects with 7,595 pairs of years with DMARD use in the first of 2 consecutive years; in 423 pairs (5.6%), subjects stopped DMARD use by year 2. In models that adjusted for RA-related factors, sociodemographics, and comorbidities, significant predictors of starting DMARDs included younger age, Hispanic ethnicity, shorter disease duration, and the use of oral glucocorticoids. In separate adjusted models, predictors of stopping DMARDs included Hispanic ethnicity and low income, while younger age was associated with a reduced risk of stopping.ConclusionEfforts to improve DMARD use should focus on patient age, ethnicity, and income and RA-related factors.

ObjectiveTo analyze the effect of sociodemographic, disease, and health system characteristics and contextual features about the community of residence on the subsequent initiation of treatment with biologic agents for rheumatoid arthritis (RA).MethodsWe analyzed data from the University of California, San Francisco Rheumatoid Arthritis Panel Study for the years 1999-2011. Principal data collection was by a structured annual phone survey. We estimated Kaplan-Meier curves of the time until initiation of biologic agents, stratified by age and income. We also used Cox regression to estimate the effect of individual-level sociodemographic and medical factors, contextual-level socioeconomic status measures, and density of health providers in the local community on the probability of initiating therapy with biologic agents for RA.ResultsIn total, 527 persons were included in the panel in 1999, and 229 persons (44%) had initiated therapy with biologic agents by 2011. In multivariable Cox regression models, age

BackgroundInsured adults receive invasive cardiovascular procedures more frequently than uninsured adults. We examined the impact of healthcare reform in Massachusetts on use of coronary revascularization procedures and in-hospital and 1-year mortality by race/ethnicity, education, and sex.Methods and resultsUsing hospital claims data, we compared differences in coronary revascularization rates (coronary artery bypass grafting or percutaneous coronary intervention) and in-hospital mortality by race/ethnicity, education, and sex among Massachusetts residents aged 21 to 64 years hospitalized with a principal discharge diagnosis of ischemic heart disease before (November 1, 2004, to July 31, 2006) and after (December 1, 2006, to September 30, 2008) reform; 1-year mortality was calculated for those undergoing revascularization. Adjusted logistic regression assessed 24 216 discharges before reform and 20 721 discharges after reform. Blacks had 30% lower odds of receiving coronary revascularization than whites in the prereform period. Compared with whites in the postreform period, blacks (odds ratio=0.73; 95% confidence interval, 0.63-0.84) and Hispanics (odds ratio= 0.84; 95% confidence interval, 0.74-0.97) were less likely and Asians (odds ratio=1.29; 95% confidence interval, 1.01-1.65) were more likely to receive coronary revascularization. Patients living in more educated communities, men, and persons with private insurance were more likely to receive coronary revascularization before and after reform. Compared with the prereform period, the adjusted odds of in-hospital mortality were higher in patients living in less-educated communities in the postreform period. No differences in 1-year mortality by race/ethnicity, education, or sex for revascularized patients were observed before or after reform.ConclusionsReducing insurance barriers to receipt of coronary revascularization procedures has not yet eliminated preexisting demographic and educational disparities in access to these procedures.

BackgroundPatient panel management and community-based care management may be viable strategies for community health centers to improve the quality of diabetes care for vulnerable patient populations. The objective of our study was to clarify implementation processes and experiences of integrating office-based medical assistant (MA) panel management and community health worker (CHW) community-based management into routine care for diabetic patients.MethodsMixed methods study with interviews and surveys of clinicians and staff participating in a study comparing the effectiveness of MA and CHW health coaching for improving diabetes care. Participants included 24 key informants in five role categories and 249 clinicians and staff survey respondents from 14 participating practices. We conducted thematic analyses of key informant interview transcripts to clarify implementation processes and describe barriers to integrating the new roles into practice. We surveyed clinicians and staff to assess differences in practice culture among intervention and control groups. We triangulated findings to identify concordant and disparate results across data sources.ResultsImplementation processes and experiences varied considerably among the practices implementing CHW and MA team-based approaches, resulting in differences in the organization of health coaching and self-management support activities. Importantly, CHW and MA responsibilities converged over time to focus on health coaching of diabetic patients. MA health coaches experienced difficulty in allocating dedicated time due to other MA responsibilities that often crowded out time for diabetic patient health coaching. Time constraints also limited the personal introduction of patients to health coaches by clinicians. Participants highlighted the importance of a supportive team climate and proactive leadership as important enablers for MAs and CHWs to implement their health coaching responsibilities and also promoted professional growth.ConclusionImplementation of team-based strategies to improve diabetes care for vulnerable populations was diverse, however all practices converged in their foci on health coaching roles of CHWs and MAs. Our study suggests that a flexible approach to implementing health coaching is more important than fidelity to rigid models that do not allow for variable allocation of responsibilities across team members. Clinicians play an instrumental role in supporting health coaches to grow into their new patient care responsibilities.

PurposeYear-to-year decreases in smoking in the US have been observed only sporadically in recent years, which suggest a need for intensified efforts to identify those at risk for persistent smoking. To address this need, we examined the association between a variety of psychosocial stressors and smoking persistence, cessation, and relapse over 9-10 years among adults in the United States (n = 4,938, ages 25-74).MethodsUsing information provided at baseline and follow-up, participants were categorized as non-smokers, persistent smokers, ex-smokers, and relapsed smokers. Stressors related to relationships, finances, work-family conflict, perceived inequality, neighborhood, discrimination, and past-year family problems were assessed at baseline and follow-up.ResultsHigh stress at both assessments was associated with greater odds of persistent smoking for stressors related to relationships, finances, work, perceived inequality, past-year family problems, and a summary score. Among respondents who were smokers at baseline, high stress at both time points for relationship stress, perceived inequality, and past-year family problems was associated with nearly double the odds of failure to quit.ConclusionsInterventions to address psychosocial stress may be important components within smoking cessation efforts.

PurposeNew strategies for delivering cancer follow-up care are needed. We surveyed primary care providers (PCPs) and oncologists to assess how physician attitudes toward and self-efficacy with cancer follow-up affect preferences for different cancer survivorship models.MethodsThe survey of physician attitudes regarding the care of cancer survivors was mailed to a randomly selected national sample of PCPs and oncologists to evaluate their perspectives regarding physician roles, knowledge about survivorship care processes, and views on cancer surveillance. Multinomial logistic regression models were constructed to examine how physician attitudes towards, and self-efficacy with, their own skills affected preferences for different cancer survivorship care models.ResultsOf 3,434 physicians identified, a total of 2,026 participants provided eligible responses: 938 PCPs and 1,088 oncologists. Most PCPs (51 %) supported a PCP/shared care model; whereas, the majority of specialists (59 %) strongly endorsed an oncologist-based model (p

Abstract Introduction Numerous studies across different health systems have documented that many patients with rheumatoid arthritis (RA) do not receive disease-modifying anti-rheumatic drugs (DMARDs). Relatively little is known about correlates of DMARD use and whether there are socioeconomic and demographic disparities. We examined DMARD use during 2001 to 2006 in the Medicare Current Beneficiary Survey (MCBS), a longitudinal US survey of randomly selected Medicare beneficiaries. Methods Participants in MCBS with RA were included in the analyses, and DMARD use was based on an in-home assessment of all medications. Variables included as potential correlates of DMARD use in weighted regression models included race/ethnicity, insurance, income, education, rheumatology visit, region, age, gender, comorbidity index, and calendar year. Results The cohort consisted of 509 MCBS participants with a diagnosis code for RA. Their median age was 70 years, 72% were female, and 24% saw a rheumatologist. Rates of DMARD use ranged from 37% among those

Abstract Background We describe early dissemination patterns for first-line bevacizumab given for metastatic colorectal cancer treatment. Methods We analyzed patient surveys and medical records for a population-based cohort with metastatic colorectal cancer treated in multiple regions and health systems in the United States (US). Eligible patients were diagnosed with metastatic colorectal cancer and initiated first-line chemotherapy after US Food & Drug Administration (FDA) bevacizumab approval in February 2004. First-line bevacizumab therapy was defined as receiving bevacizumab within 8 weeks of starting chemotherapy for metastatic colorectal cancer. We evaluated factors associated with first-line bevacizumab treatment using logistic regression. Results Among 355 patients, 31% received first-line bevacizumab in the two years after FDA approval, including 26% of men, 41% of women, and 16% of those ≥ 75 years. Use rose sharply within 6 months after FDA approval, then plateaued. 20% of patients received bevacizumab in combination with irinotecan; 53% received it with oxaliplatin. Men were less likely than women to receive bevacizumab (adjusted OR 0.55; 95% CI 0.32-0.93; p = 0.026). Patients ≥ 75 years were less likely to receive bevacizumab than patients < 55 years (adjusted OR 0.13; 95% CI 0.04-0.46; p = 0.001). Conclusions One-third of eligible metastatic colorectal cancer patients received first-line bevacizumab shortly after FDA approval. Most patients did not receive bevacizumab as part of the regimen used in the pivotal study leading to FDA approval.