Jason A. Thompson

Health Literacy, Cognitive Abilities, and Mortality Among Elderly Persons

BackgroundLow health literacy and low cognitive abilities both predict mortality, but no study has jointly examined these relationships.MethodsWe conducted a prospective cohort study of 3,260 community-dwelling adults age 65 and older. Participants were interviewed in 1997 and administered the Short Test of Functional Health Literacy in Adults and the Mini Mental Status Examination. Mortality was determined using the National Death Index through 2003.Measurements and Main ResultsIn multivariate models with only literacy (not cognition), the adjusted hazard ratio was 1.50 (95% confidence of interval [CI] 1.24–1.81) for inadequate versus adequate literacy. In multivariate models without literacy, delayed recall of 3 items and the ability to serial subtract numbers were associated with higher mortality (e.g., adjusted hazard ratios [AHR] 1.74 [95% CI 1.30–2.34] for recall of zero versus 3 items, and 1.32 [95% CI 1.09–1.60] for 0–2 vs 5 correct subtractions). In multivariate analysis with both literacy and cognition, the AHRs for the cognition items were similar, but the AHR for inadequate literacy decreased to 1.27 (95% CI 1.03 – 1.57).ConclusionsBoth health literacy and cognitive abilities independently predict mortality. Interventions to improve patient knowledge and self-management skills should consider both the reading level and cognitive demands of the materials.

Disparities in Enrollment and Use of an Electronic Patient Portal

BackgroundWith emphasis on the meaningful use of electronic health records, patient portals are likely to become increasingly important. Little is known about patient enrollment in, and use of, patient portals after explicit invitation from providers.ObjectivesTo examine enrollment in, and use of, an electronic patient portal by race/ethnicity, gender and age.DesignObservational, cross sectional study.ParticipantsPatients with attending physicians seen at one urban, academic primary care practice between May 2008 and October 2009 who received electronic orders inviting their participation in an electronic patient portal.Main Measures(a) Enrollment in the patient portal, (b) Solicitation of provider advice among enrollees, (c) Requests for medication refills among enrollees.Key ResultsOverall, 69% of 7,088 patients enrolled in the patient portal. All minority patients were significantly less likely to enroll than whites: 55% blacks, 64% Latinos and 66% Asians compared with 74% whites (chi-square p < 0.05 for all pairwise comparisons). These disparities persisted in adjusted analyses, although differences for Asians were no longer significant. In addition, the oldest patients were less likely to enroll than the youngest (adjusted OR 0.79, 95% CI 0.65–0.97). Although there were no racial/ethnic disparities in use of the patient portal among enrollees, we found differences by age and gender. The youngest patients were significantly less likely to solicit provider advice or request medication refills than any other age group in unadjusted and adjusted analyses. Similarly, male patients were less likely to solicit provider advice than women in all analyses.ConclusionLarge racial/ethnic disparities were seen in enrollment in our patient portal. Among enrollees, use of the portal was similar by race/ethnicity, but not by age or gender. Future efforts to expand use of the patient portal need to address potential mechanisms for these disparities to ensure this technology is accessible to diverse patient populations.

In search of 'low health literacy': Threshold vs. gradient effect of literacy on health status and mortality

Studies have demonstrated significant associations between limited literacy and health outcomes. Yet differences in literacy measurement and the cutoffs used for analysis have made it difficult to fully understand the relationship between literacy and health across the entire spectrum of literacy (i.e., whether the relationship is continuous and graded or whether a threshold exists below which literacy is independently associated with health). To analyze this question, we re-examined the relationship between literacy, baseline physical functioning and mental health, and all-cause mortality for a cohort of 3260 US community-dwelling elderly who were interviewed in 1997 to determine demographics, socioeconomic status, chronic conditions, self-reported physical and mental health (SF-36 subscales), health behaviors, and literacy based upon the Short Test of Functional Health Literacy in Adults (S-TOFHLA). All-cause mortality was determined using data from the US National Death Index through 2003. Seven categories of S-TOFHLA literacy scores were created and used in this analysis instead of the existing three categories identified with the measure. In multivariate analyses, a continuous, graded relationship between literacy and baseline physical functioning was identified. However, participants scoring below the third literacy category had significantly worse mental health compared to the highest literacy category, displaying a notable threshold. Finally, all six literacy categories were significantly associated with greater all-cause mortality risk compared to the highest literacy category, but again there was a marked threshold below the third category at which the adjusted mortality rate significantly increased compared to all other categories. We conclude that the nature of the relationship between literacy and health may vary depending upon the outcome under examination.

Changes in Performance After Implementation of a Multifaceted Electronic-Health-Record-Based Quality Improvement System

Background:Electronic health record (EHR) systems have the potential to revolutionize quality improvement (QI) methods by enhancing quality measurement and integrating multiple proven QI strategies. Objectives:To implement and evaluate a multifaceted QI intervention using EHR tools to improve quality measurement (including capture of contraindications and patient refusals), make point-of-care reminders more accurate, and provide more valid and responsive clinician feedback (including lists of patients not receiving essential medications) for 16 chronic disease and preventive service measures. Design:Time series analysis at a large internal medicine practice using a commercial EHR. Subjects:All adult patients eligible for each measure (range approximately 100–7500). Measures:The proportion of eligible patients who satisfied each measure after removing those with exceptions from the denominator. Results:During the year before the intervention, performance improved significantly for 8 measures. During the year after the intervention, performance improved significantly for 14 measures. For 9 measures, the primary outcome improved more rapidly during the intervention year than during the previous year (P < 0.001 for 8 measures, P = 0.02 for 1). Four other measures improved at rates that were not significantly different from the previous year. Improvements resulted from increases in patients receiving the service, documentation of exceptions, or a combination of both. For 5 drug-prescribing measures, more than half of physicians achieved 100% performance. Conclusions:Implementation of a multifaceted QI intervention using EHR tools to improve quality measurement and the accuracy and timeliness of clinician feedback improved performance and/or accelerated the rate of improvement for multiple measures simultaneously.

Health insurance coverage and the risk of decline in overall health and death among the near elderly, 1992-2002.

Background:Although individuals’ health insurance coverage changes frequently, previous analyses have not accounted for changes in insurance coverage over time. Objective:We sought to determine the independent association between lack of insurance and the risk of a decline in self-reported overall health and death from 1992 to 2002, accounting for changes in self-reported overall health and insurance coverage. Methods:We analyzed data from the Health and Retirement study, a prospective cohort study of a national sample of community-dwelling adults age 51–61 years old at baseline. Major decline in self-reported overall health and mortality was determined at 2-year intervals. Results:People who were uninsured at baseline had a 35% (95% confidence interval [CI] 12–62%) higher risk-adjusted mortality from 1992 to 2002 compared with those with private insurance. However, when we analyzed outcomes over 2-year intervals, individuals who were uninsured at the start of each interval were more likely to have a major decline in their overall health (pooled adjusted relative risk 1.43, 95% CI 1.28–1.63), but they were equally likely to die (pooled adjusted relative risk 0.96, 95% CI 0.73–1.27). Of the 1512 people who were uninsured at baseline, 220 (14.6%) died; of those who died, only 70 (31.8%) were still uninsured at the HRS interview prior to death. Conclusions:Death does not appear to be a short-term consequence of being uninsured. Instead, higher long-term mortality among the uninsured results from erosion in this populations health status over time and the attendant higher mortality associated with this. Most deaths among the uninsured occur after individuals have gained either public or private health insurance.

A census-based analysis of racial disparities in lower extremity amputation rates in Northern Illinois, 1987-2004

BACKGROUND Given improvements in care for peripheral vascular disease and diabetes over the last two decades, it was of interest whether racial disparities in lower extremity amputation rates had changed. METHODS Hospital data for 18 years (1987-2004) were used to compute above, below, and through foot amputation rates for over eight million people living in the Chicago metropolitan area. Three areas were created from zip code level census data. Differences in amputation rates were compared between residents of zip code areas that were >50% African American, 10% to 50% African American, or <10% African American. RESULTS The largely African American area of the South and West sides of Chicago, with less than 15% of the area population, accounted for 27% of all amputation discharges (n = 33,775) over the 18 years. For all residents of northern Illinois, major (above and below knee) amputation rates declined to 17 per 100,000 residents over the last decade, and both inpatient mortality and length of stay fell throughout the period. However, residents of largely African American zip codes had over five times higher per capita amputation rates than residents of primarily white zip codes. CONCLUSIONS Racial disparities have remained remarkably constant, despite progress in reducing the overall major amputation rate in northern Illinois. Addressing these disparities will require that low income, medically complex patients at risk of limb loss receive timelier, high performance care, combined with community-based public health and preventive medicine interventions that address the social determinants of health.

America's Best Medical Schools: A Critique of the U.S. News & World Report Rankings

Rankings of American medical schools published annually by the news magazine U.S. News & World Report are widely used to judge the quality of the schools and their programs. The authors describe and then critique the rankings on methodologic and conceptual grounds, arguing that the annual U.S. News medical school evaluation falls short in both areas. Three categories of program quality indicators different from those used by U.S. News are presented as alternative ways to judge medical schools. The authors conclude that the annual U.S. News & World Report rankings of American medical schools are ill-conceived; are unscientific; are conducted poorly; ignore medical school accreditation; judge medical school quality from a narrow, elitist perspective; and do not consider social and professional outcomes in program quality calculations. The medical school rankings have no practical value and fail to meet standards of journalistic ethics.

Patients’ attitudes toward health care providers collecting information about their race and ethnicity

AbstractBACKGROUND: Experts recommend that health care providers (HCPs) collect patients’ race/ethnicity, but HCPs worry that this may alienate patients. OBJECTIVE: To determine patients’ attitudes toward HCPs collecting race/ethnicity data. DESIGN: Cross-sectional survey. PARTICIPANTS: General Internal Medicine patients (n = 220). MEASUREMENTS: Perceived importance of having HCPs collect race/ethnicity data, their concerns about this, comfort level providing this information, and reactions to 4 statements explaining the rationale for collecting this. RESULTS: Approximately 80% somewhat or strongly agreed that HCPs should collect information on patients’ race/ethnicity. However, 28% had significant discomfort (score 5 or less on 10-point scale) reporting their own race/ethnicity to a clerk, and 58% were somewhat or very concerned that this information could be used to discriminate against patients. Compared with whites, blacks, and Hispanics felt less strongly that HCPs should collect race/ethnicity data from patients (P=.04 for both pairwise comparisons), and blacks were less comfortable reporting their own race/ethnicity than whites (P=.03). Telling patients that this information would be used for monitoring quality of care improved comfort more than telling patients that the data collected (a) was mandated by others, (b) would be used to guide staff hiring and training, and (c) would be used to ensure the patient got the best care possible. CONCLUSIONS: Most patients think HCPs should collect information about race/ethnicity, but many feel uncomfortable giving this information, especially among minorities. Health care providers can increase patients’ comfort levels by telling them this will be used to monitor quality of care.

Medication Dosing Errors for Patients with Renal Insufficiency in Ambulatory Care

BACKGROUND Approximately 8.7 million (3%) Americans have various degrees of chronic renal insufficiency (CRI). Patients with CRI are susceptible to adverse drug events related to improper dose adjustment of drugs that are eliminated primarily unchanged through the kidney. Renal dosing errors are an important quality-of-care problem in the inpatient setting, yet little is known about dosing errors for patients with renal insufficiency in the outpatient setting. METHODS Electronic records were queried to identify patients with CRI (estimated creatinine clearance < 50 mL/min) who visited the ambulatory care clinic at least once from January 1, 2003 through December 31, 2003. RESULTS Of the total of 224 patients identified with CRI, 157 (70%) received one or more of 17 drugs with high rates of renal elimination. A total of 207 drugs requiring dose adjustment were prescribed to these patients, and 52 (25%) were prescribed at an inappropriately high dose. For 127 (57%) of the 224 patients, CRI was not documented. Patients with documented CRI were equally likely to be prescribed an inappropriately high dose of a target drug. DISCUSSION Incorrect dosing of medications among patients with CRI is common in the ambulatory care setting. Strategies for preventing medication dosing errors can target the prescribing and monitoring stages of pharmaceutical care.

Development of a measure of attitude toward nutrition in patient care

BACKGROUND Development of reliable measures of medical student and resident attitudes about nutrition in patient care is needed before the effects of educational interventions or clinical experience can be gauged. This report describes the systematic development of a measure of attitude toward nutrition in patient care. It presents evidence about scale reliability and the absence of response bias that endorses the trustworthiness of data from the measure. METHODS An eight-step attitude scale development procedure was used to create the Nutrition In Patient care Survey (NIPS). Data from five samples of first- and second-year medical students and first-year medical residents were subjected to factor analysis (PA2, varimax rotation), reliability analyses, and statistical analyses to test for demographic bias in the attitude data. RESULTS A 45-item attitude measure was developed that contains five subscales derived from the factor analysis: (1) nutrition in routine care (NRC, 8 items); (2) clinical behavior (CB, 20 items); (3) physician-patient relationship (PPR, 8 items); (4) patient behavior/motivation (PBM, 3 items); and (5) physician efficacy (PE, 6 items). Each subscale yields reliable data in terms of internal consistency (alpha coefficients) and stability (test-retest reliability). Medical student and resident demographic variables have negligible influence on attitude scores. DISCUSSION The NIPS subscales yield reliable data that can be used to assess outcomes in evaluation research on educational or clinical interventions or to predict patient care practices. Systematic attitude scale development increases the likelihood that the resulting measures will produce useful, trustworthy data.