******************** START MESSAGE *****************
year
second
title
ELEVATED INCIDENCE OF DIAGNOSED CREUTZFELDT-JAKOB DISEASE: VETERANS HEALTH ADMINISTRATION SYSTEM, FISCAL YEARS 1997-1998
abstract
Purpose Creutzfeldt-Jakob disease (CJD) is a rare and rapidly progressive dementing illness. Though most cases of this prion disease have no known etiology, CJD may be acquired through infectious transmission, as evidenced epidemiologically by temporal and spatial clustering around a common exposure. Military service and subsequent medical care are related factors that have yet to be fully explored in this context. Therefore, we sought to: 1) describe the annual incidence of CJD diagnosis in the national Veterans Health Administration (VHA) medical system from fiscal years (FY) 1997-2010, 2) to identify “spikes” in incidence that could potentially reflect infectious clustering mechanisms, and 3) to provide context for these findings through comparison with a standardized American population. Methods We first analyzed the VHA Medical SAS Datasets to identify incident cases with a specified ICD-9 code for CJD (FY 1997-2008: 046.1, FY 2009-2010: 046.11 and 046.19). !
Electronic medical records for all cases were then accessed through VISTA Web in order to determine whether the available data indicated a diagnosis of CJD, an alternative diagnosis (where CJD was considered), or miscoding (where CJD was not considered). Whenever possible, documentation of history, physical examination, electroencephalography, magnetic resonance imaging, cerebrospinal fluid evaluation, and neuropathology was evaluated to verify if cases met diagnostic criteria for CJD (possible, probable, or definite). Cases coded as and/or clinically suspected to be CJD but with insufficient available data to meet these criteria were considered “suspect.” The annual incidence of CJD diagnoses from all categories was then calculated as a function of the number of unique patients seen in treatment in the VA each year. Poisson point estimates were used to determine whether the annual incidence of CJD diagnosis in any consecutive two-year period was significantly greater than any of the others. The annual standardized mortality ratio (SMR), based on the American population from 1989-1998 adjusted for age and sex, was determined to offer context for these findings outside the VHA system. Results Overall, there were 115 cases of CJD diagnosed in the VHA between FY 1997 and 2010, for an incidence of 2.0 cases per million per year (range: 0.8-3.7 cases per million annually). No cases appeared consistent with variant CJD. An elevated incidence of 3.6 per million during FY 1997-1998 was greater than any other consecutive two-year period in the VHA, meeting significance for all other periods except FY 1998-1999. The annual SMR (VHA/U.S.) was less than or equal to 1 for all years, with the exception of FY 1997 and FY 1998, which were 1.47 and 1.32, respectively.
Impact The incidence of CJD diagnosis in the VHA system during FY 1997-1998 was greater than that in any other consecutive two-year stretch from FY 1997 to 2010 and the only one greater than expected in an American population standardized for age and sex during that same time. The cause of this spike in diagnoses remains unexplained and deserves further investigation into the possibility of infectious clustering.
Key Words: Creutzfeldt-Jakob disease, Incidence, Veterans Health Adminstration
additional_authors
Jed A. Barash, M.D., Robert Wood Johnson Clinical Scholars Program, Yale University School of Medicine; Rani A. Desai, Ph.D., M.P.H., Department of Epidemiology and Public Health, Yale University School of Public Health; Huned S. Patwa, M.D., Department of Neurology, Yale University School of Medicine. All authors are jointly affiliated with Veteran Affairs Medical Center, West Haven, CT.
fname
Jed
lname
Barash
affiliation
Yale University (VA)
email

******************** END OF MESSAGE *****************
******************** START MESSAGE *****************
year
second
title
GENDER AND THE RISK OF HOMELESSNESS AMONG IRAQ/AFGHANISTAN VETERANS
abstract
Purpose: Veterans are disproportionately represented among homeless persons. Homelessness is associated with significant unmet health care needs as well as increased morbidity and mortality. Consequently, the Veterans Health Administration (VHA) provides housing, job training, and health care assistance to homeless Veterans through specialized homeless services programs. The magnitude of homelessness among returning Iraq/Afghanistan Veterans, a Veteran cohort with the largest proportion of women to serve and to be exposed to combat, is unknown. Therefore, we sought 1) to determine the risk of homelessness among Iraq/Afghanistan Veterans in VHA care, 2) whether this risk differs between female and male Veterans, and 3) to compare the characteristics of homeless female and male Veterans. Methods: Using the Operation Enduring Freedom/Operation Iraqi Freedom Roster, a database of Iraq/Afghanistan military service personnel enrolled in VHA care, we identified Veterans with at least one inpatient or outpatient VHA visit from October 1, 2001 to September 30, 2009. We operationalized homelessness as use of a VHA homeless services program. The study’s primary outcome, time to first use of a VHA homeless services program, was defined as the time from the date of military separation to first use of a VHA homeless services program. Cox proportional hazards regression was used to estimate the relative risk of homelessness, adjusting for relevant socio-demographic and clinical variables such as mental health conditions. We used chi-square and Wilcoxon rank-sum tests to compare the characteristics of homeless female and male Veterans.
Results: Of 445,319 Veterans, 7,431 (1.7%) met our study’s definition of homelessness, 961 women (1.8%) and 6,470 men (1.7%), during a median duration of study follow-up of 1172 days (interquartile range, 636-1651). Women were as likely as men to become homeless (adjusted hazard ratio, 1.02, 95% confidence interval, 0.95-1.09). Median time to homelessness was similar for female and male Veterans (685 days vs. 687 days, respectively, p>0.05). A larger proportion of homeless female Veterans were Black (46.9% vs. 25.6%, p<0.001), unmarried (73.4% vs. 66.4%, p<0.001), and had more than a high school education (14.2% vs. 9.0%, p<0.001) compared to homeless male Veterans. While a greater percentage of homeless women had a diagnosis of major depression (20.4% vs. 14.0%, p<0.001), homeless women had lower rates of alcohol use disorder (14.9% vs. 27.7%, p <0.001), substance use disorder (10.9% vs. 21.2%, p<0.001), PTSD (35.7% vs. 42.6%, p<0.001), and schizophrenia (0.9% vs. 1.9%, p<0.05) than homeless men. Conclusions: About 1.7% of Iraq/Afghanistan Veterans in VHA care became homeless during the study follow-up period. Overall, there was no substantial difference in the risk of homelessness between women and men. However, homeless female and male Veterans differed significantly with respect to socio-demographic and clinical characteristics. In light of our study findings, we recommend health care providers, homeless services programs, and homelessness prevention efforts address the unique and shared needs of female and male Iraq/Afghanistan Veterans.
Keywords: women, female, Veterans, homelessness, housing
additional_authors
Oni Blackstock, MD, Yale School of Medicine/VA Connecticut Healthcare System; Sally Haskell, MD, Yale School of Medicine/VA Connecticut Healthcare System; Cynthia Brandt, MD, MPH, Yale School of Medicine/VA Connecticut Healthcare System; Rani Desai, PhD, MPH, Yale School of Medicine/VA Connecticut Healthcare System;
fname
Oni
lname
Blackstock
affiliation
Yale School of Medicine/VA Connecticut Healthcare System
email

******************** END OF MESSAGE *****************
******************** START MESSAGE *****************
year
second
title
Does MRI Utilization Reduce Recurrent Stroke?
abstract
Purpose: Magnetic Resonance Imaging (MRI) is commonly as part of the acute stroke evaluation, although there is little evidence that MRI effects meaningful outcomes. We sought to determine whether receiving an MRI during stroke hospitalization reduces the risk of recurrent stroke. Methods: We identified all hospital discharges with a primary ICD-9-CM discharge diagnosis of ischemic stroke (433.x1, 434.x1, 436) or transient ischemic attack (TIA) (435) admitted through the Emergency Department in 6 states (NY, FL, NE, NC, AZ, WA). Data was drawn from the State Inpatient Databases (SID) between 2003-2009 and hospital level covariates were obtained from Medicare cost-reports. MRI utilization was identified using the SID Magnetic Resonance Technology (MRT) definition relying on UB-92 revenue codes (MRI 610-619) and ICD-9-CM procedure codes. Our primary outcome is recurrent stroke at one year, defined as any rehospitalization with a primary ischemic stroke ICD-9-CM diagnosis code. We used regional MRI practice patterns as an instrumental variable to estimate the association between MRI and recurrent stroke, defined as the number of MRIs per 1000 hospital discharges aggregated at the hospital referral region (HRR) and divided into quintiles. A 2-state probit model was used to estimate the IV effect, adjusting for patient-level (demographics, vascular risk factors, charlson comorbidities, socioeconomic status) and hospital-level (bed size, academic affiliation, volume of patients with stroke, non-stroke rehospitalization rate, adjusted mortality) variables. Results: The index population was 433, 073 ischemic stroke or TIA discharges. Mean age was 72 (SD 14) and 55.5% of the patient population was female. Recurrent stroke was indentified in 5.5% of discharges by 90 days and 7.9% of discharges by one year. Our instrument met statistical criteria for validity. Regional and individual MRI use were strongly associated — the MRI utilization rate was 36.2% in the lowest quintile and 66.7% in the highest quintile. Qualitatively, patient-level covariates were relatively evenly distributed among regional MRI utilization quintiles. Unadjusted one-year stroke recurrence rates varied little over regional MRI utilization quintiles, from 7.8% in the lowest quintile to 7.6% in the highest quintile. Our preliminary IV analysis demonstrated that patients with a higher probability of receiving MRI were no less likely to experience a recurrent stroke within 90 days (probit coefficient
sizeboxw
486
additional_authors
1. James F Burke MD, Department of Veterans Affairs, VA Center for Clinical Management and Research, Robert Wood Johnson Clinical Scholars Program and Department of Neurology, University of Michigan 2. Lesli E Skolarus MD MS, Department of Neurology, University of Michigan 3. Kevin A Kerber MD MD, Department of Neurology, University of Michigan 4. Colin R Cooke MD MS, Division of Pulmonary and Critical Care Department of Medicine, University of Michigan 5. Rodney A Hayward MD, Department of Veterans Affairs, VA Center for Clinical Management and Research, Robert Wood Johnson Clinical Scholars Program and Department of Internal Medicine, University of Michigan 6. Lewis B Morgenstern MD, Department of Neurology, University of Michigan
fname
James
lname
Burke
affiliation
University of Michigan
email

******************** END OF MESSAGE *****************
******************** START MESSAGE *****************
year
second
title
DO IMPROVEMENTS IN RACE/ETHNICITY DATA COLLECTION AFFECT KNOWLEDGE OF HEALTHCARE DISPARITIES? THE CASE OF ASIAN AMERICANS AND ACUTE MYOCARDIAL INFARCTION
abstract
Background: Studies of healthcare disparities among racial/ethnic minorities are significantly limited by the use of broad race categories to identify individuals from diverse backgrounds and inconsistent methods of data collection for race/ethnicity. Recognizing these limitations, the Institute of Medicine recommends the collection of granular ethnicity data (defined as a “person’s ethnic origin or descent, ‘roots,’ or heritage, or the place of birth of the person or the person’s parents or ancestors”) and calls for research to identify more effective methods of collecting self-reported race/ethnicity data. Healthcare institutions are implementing these recommendations to varying degrees based on organizational capacity. Since 2005, New Jersey hospitals have collected and reported granular ethnicity data for the 6 largest Asian-American subgroups in the state. In 2007, the Health Research and Education Trust of New Jersey disseminated a toolkit to systematically improve the accuracy and reliability of patient race/ethnicity data collected by hospitals. However, the impact of efforts to improve race/ethnicity data collection on knowledge of healthcare disparities is unknown.
Purpose: To explore the effect of a statewide strategy to improve the quality of race/ethnicity data collection and reporting on hospitalization rates for acute myocardial infarction (AMI) among Asian-American subgroups.
Methods: We conducted a cross-sectional comparison of hospitalization rates for AMI among Asian Americans before (2005-2006) and after (2008-2009) the implementation of the statewide strategy to improve race/ethnicity data collection using the New Jersey State Inpatient Databases (SID). Rates were compared for Asian Indian, Chinese, Filipino, Japanese, Korean, and Vietnamese subgroups. White, black, and Asian racial groups were included as reference categories. We used the direct standardization method to calculate age-adjusted AMI hosp!
italization rates among adults age 35 and over from each racial/ethnic group. AMI hospitalizations were identified by ICD-9 diagnostic codes from the SID and used as the numerator in rate calculations. Population estimates for specific Asian-American subgroups and race groups from the U.S. Census Bureau’s annual American Community Survey were used as the denominator. Rates were age-adjusted to the 2000 U.S. standard population. 95% confidence intervals for age-adjusted rates will be calculated. Results: AMI hospitalization rates decreased for whites (429 versus 381), blacks (364 versus 359), and Vietnamese (159 versus 126) after implementation of the toolkit to improve race/ethnicity data collection. The rates increased for the overall Asian race category (159 versus 228), Asian Indians (141 versus 214), Chinese (58 versus 98), Filipino (145 versus 174), Japanese (131 versus 291), and Koreans (73 versus 164) after implementation of the toolkit.
Conclusions: Implementation of a standardized and systematic strategy to coll!
ect race/ethnicity data in hospital settings revealed higher rates of hospitalization for AMI among all Asian-American subgroups except Vietnamese. Our findings suggest that the use of a systematic approach to the collection of race/ethnicity data, particularly granular ethnicity data, may improve our knowledge of healthcare disparities among specific minority populations. Key Words: Healthcare Disparities, Data Collection, Asian Americans, Myocardial Infarction
additional_authors
Rosette J. Chakkalakal, MD, Robert Wood Johnson Clinical Scholars Program, Department of Medicine, Yale University School of Medicine; Mayur M. Desai, MPH, PhD, Division of Chronic Disease Epidemiology, Yale University School of Public Health; Brahmajee K. Nallamothu, MD, MPH, Division of Cardiovascular Medicine and Center for Healthcare Outcomes and Policy, University of Michigan Medical School; Harlan M. Krumholz, MD, SM, Section of Cardiovascular Medicine and the Robert Wood Johnson Clinical Scholars Program, Department of Medicine, Section of Health Policy and Administration, Department of Epidemiology and Public Health, Yale University School of Medicine, Center for Outcomes Research and Evaluation, Yale New Haven Hospital; Marcella Nunez-Smith, MD, Robert Wood Johnson Clinical Scholars Program and Section of General Medicine, Department of Medicine, Yale University School of Medicine; Justin P. Fox, MD, Robert Wood Johnson Clinical Scholars Program, Department of Medicine, Yale University School of Medicine; Nanlesta R. Pilgrim, PhD, Johns Hopkins Bloomberg School of Public Health; Jacinda Dariotis, PhD, Johns Hopkins Bloomberg School of Public Health; Romana Hasnain-Wynia, PhD, Center for Healthcare Equity, Institute for Healthcare Studies, Northwestern University, Feinberg School of Medicine

fname
Rosette
lname
Chakkalakal
affiliation
Yale University School of Medicine
email

******************** END OF MESSAGE *****************
******************** START MESSAGE *****************
year
second
title
MEDICAL HOME RECOGNITION AND DIABETES QUALITY OF CARE IN COMMUNITY HEALTH CENTERS
abstract
Medical home, community health centers Test whether NCQA level of medical home recognition is associated with quality of diabetes care provided by community health centers (CHCs). We partnered with urban safety net CHCs from the Community Clinic Association of Los Angeles County (CCALAC). Of the CCALAC's 40 member organizations, 27 (68%) participated. In a previous phase, these 27 organizations completed the NCQA’s medical home survey. The NCQA characterizes a practice into four levels from not recognized as a medical home to Level III based on a 100-point scale and number of “must-pass” elements passed. We assigned each clinic a level based on these criteria. In this phase of the study, we collected patient demographic information and diabetes care processes and intermediate outcomes for 50 randomly identified diabetic patients from each participating clinic. We collected clinical measures related to glucose, cholesterol, and blood pressure control and appropriate screening tests. We used thresholds to create dichotomous outcomes: whether screening tests were conducted in the past 12 months or whether a risk factor was adequately controlled. We used a random intercept mixed effects model adjusting for patient- and clinic-level characteristics and clustering within clinics. We created odds ratios for each diabetes outcome comparing clinics at higher levels of NCQA recognition to those at the lowest level. Each of the 27 participating CHCs exceeded the minimum threshold and would be recognized by the NCQA as a medical home. Seven (26%) earned the highest Level III recognition, two (7%) received Level II, and 18 (67%) received Level I. Of these 18 at Level I, 14 scored enough points to earn a Level II or III recognition, but automatically received a Level I designation because they failed at least one of the 10 must-pass elements. We therefore created three levels of comparison: 1) Level I by points as the lowest level and the reference; 2) Level I by missed must-pass elements clinics as the intermediate; and 3) Level II and III clinics combined because of Level II’s small sample size as the highest. We dropped 45 observations because of missing values creating a final sample of 1305 diabetic patients clustered within 27 organizations. The odds ratios comparing the intermediate and highest NCQA level clinics to the lowest level clinics were non-significant for all diabetes outcomes. These initial analyses indicate that performance on the NCQA medical home recognition instrument is not associated with diabetes quality of care for the participating urban safety net clinics. The odds ratios from the multi-level model demonstrate that a diabetic patient has a similar likelihood of receiving a screening test or having a risk factor controlled in clinics with the highest and lowest NCQA medical home recognition. Our findings raise the possibility that the NCQA tool may not be the best mechanism for evaluating quality of diabetes care in CHCs. The implications of this are significant as HRSA and CMS have both selected the NCQA’s survey as their preferred medical home assessment for CHCs.
additional_authors
Robin Clarke, M.D., Robert Wood Johnson Clinical Scholars Program, UCLA Chi-hong Tseng, Ph.D, Division of General Internal Medicine and Health Services Research, UCLA David Geffen School of Medicine Brian Mittman, Ph.D, VA Center for Implementation Practice and Research Support Arleen Brown, M.D., Ph.D, Division of General Internal Medicine and Health Services Research, UCLA David Geffen School of Medicine
fname
Robin
lname
Clarke
affiliation
UCLA
email

******************** END OF MESSAGE *****************
******************** START MESSAGE *****************
year
second
title
PANDEMIC PLANNING IN PEDIATRIC EMERGENCY DEPARTMENTS: SURGE EXPERIENCES DURING THE 2009 H1N1 INFLUENZA PANDEMIC
abstract
In the United States, terrorist attacks, natural disasters, and pandemics of the past decade have prompted hospitals to better plan and prepare for large scale public health emergencies. A central goal in preparing for all major medical events is to optimize medical surge capacity - the ability to evaluate and care for a markedly increased volume of patients. Recently, pediatric emergency departments (EDs) experienced medical surge during the 2009 H1N1 influenza pandemic. Little is known about how pandemic planning efforts prepared pediatric EDs to manage medical surge in this context. The objective of this study is to characterize pediatric ED providers’ experiences with pandemic planning and implementation during the 2009 H1N1 pandemic. From May through August 2011, the authors are conducting a qualitative study consisting of in-depth phone interviews with pediatric ED medical directors or designated respondents with expertise in ED pandemic planning. Purposive sampling is being used to select participants from institutions with pediatric emergency medicine training programs across the United States representing a range of geographic regions, hospital types and annual ED volumes. Data are being collected through use of a standardized interview guide. Prompts explore respondents’ experiences with development and implementation of their respective institution’s pandemic plan, challenges faced and solutions adopted in response to medical surge, and communication with the public and emergency preparedness practitioners. Interviews are being audio-recorded and professionally transcribed. Our analysis will be guided by the constant comparative method in order to develop unifying themes characterizing how pandemic plans were developed and implemented in pediatric EDs during the H1N1 influenza pandemic of 2009. We will describe the pandemic planning and implementation experiences of the cohort as well as areas in which respondents perceived they were not adequately prepared, and where they found their plans to be successful during the 2009 H1N1 influenza pandemic. We anticipate our results will reveal relationships between pandemic planning efforts and perceived experiences in managing pediatric ED surge associated with the pandemic. Understanding the impact of pandemic planning on management of pandemic-associated pediatric ED surge may help practitioners and policymakers better prepare for future pandemics or disaster events associated with sustained ED surge.
additional_authors
Clara Filice, MD, MPH, Robert Wood Johnson Foundation Clinical Scholars Program, Yale University School of Medicine; Federico Vaca, MD, Emergency Medicine Department, Yale University School of Medicine; Leslie Curry, PhD, Robert Wood Johnson Foundation Clinical Scholars Program; Stephanie Platis, Yale School of Public Health; Nicole Lurie, MD, MPH, HHS Assistant Secretary for Preparedness and Response; Sandy Bogucki, MD, PhD, Emergency Medicine Department, Yale University School of Medicine
fname
Clara
lname
Filice
affiliation
Yale University
email

******************** END OF MESSAGE *****************
******************** START MESSAGE *****************
The Police Health Survey: Community-Partnered Research with the New Haven Police Department