Search

Background: Human papillomavirus (HPV) vaccination rates in the United States remain low and lag behind other recommended adolescent vaccines. Studies evaluating the association of geographic and area-level characteristics with HPV vaccination rates provide a valuable resource for public health planning.Method: We used the Rochester Epidemiology Project data linkage system to ascertain HPV vaccination rates between 2010 and 2015 in a 7-county region of southern Minnesota. Geocoded individual patient data were spatially linked to socioeconomic data from the American Community Survey at the census block group level. Bayesian hierarchical logistic regression was used to model incident vaccination rates, adjusting for individual- and area-level sociodemographic characteristics, and geolocation. Geolocation was modeled as an approximated Gaussian field using a Stochastic Partial Differential Equations approach. All models were estimated using Integrated Nested Laplace Approximations.Results: In adjusted models, increasing age and female sex were associated with increased HPV vaccination. Lower socioeconomic status was associated with decreased rates of initiation [adjusted odds ratio (AOR); 95% confidence interval = 0.90 (0.86–0.95)], completion of the second dose [AOR = 0.88 (0.83–0.93)], and completion of the third dose [AOR = 0.85 (0.80–0.92)]. Geographic spatial analysis demonstrated increased odds of vaccination for the eastern region and in the greater Rochester metropolitan area, showing significant spatial variation not explained by individual level characteristics and ACS block group-level data.Conclusions: HPV vaccination rates varied geographically and by individual and geographically indexed sociodemographic characteristics.Impact: Identifying geographic regions with low HPV vaccination rates can help target clinical and community efforts to improve vaccination rates. Cancer Epidemiol Biomarkers Prev; 26(4); 533–40. ©2017 AACR.See all the articles in this CEBP Focus section, “Geospatial Approaches to Cancer Control and Population Sciences.”

To assess the spatial variation and the association of neighborhood/area-level effects and individual characteristics we implemented a Bayesian spatial logistic regression model.

Background: The Rochester Epidemiology Project (REP) medical records-linkage system offers a unique opportunity to integrate medical and residency data with existing environmental data, to estimate individual-level exposures. Our primary aim was to provide an archetype of this integration. Our secondary aim was to explore the association between groundwater inorganic nitrogen concentration and adverse child and adolescent health outcomes. Methods: We conducted a nested case-control study in children, aged seven to eighteen, from six counties of southeastern Minnesota. Groundwater inorganic nitrogen concentration data were interpolated, to estimate exposure across our study region. Residency data were then overlaid, to estimate individual-level exposure for our entire study population (n = 29,270). Clinical classification software sets of diagnostic codes were used to determine the presence of 21 clinical conditions. Regression models were adjusted for age, sex, race, and rurality. Results: The analyses support further investigation of associations between nitrogen concentration and chronic obstructive pulmonary disease and bronchiectasis (OR: 2.38, CI: 1.64–3.46) among boys and girls, thyroid disorders (OR: 1.44, CI: 1.05–1.99) and suicide and intentional self-inflicted injury (OR: 1.37, CI: >1.00–1.87) among girls, and attention deficit conduct and disruptive behavior disorders (OR: 1.34, CI: 1.24–1.46) among boys. Conclusions: Investigators with environmental health research questions should leverage the well-enumerated population and residency data in the REP.

End-of-life care differs by serious illness diagnosis. Cancer and dementia are serious illnesses that have been associated with less intensive end-of-life health care use. It is not known how health care utilization varies in the presence of1 serious illness.We used the Rochester Epidemiology Project to identify persons living in a midwestern area who died on July 1, 2017-June 30, 2018 at age ≥65 years, and were seriously ill. We examined the number of emergency department (ED), hospital, and intensive care unit (ICU) stays in the last 6 months and the last 30 days of life. We used Poisson regression to determine the incidence rate ratio for ED, hospital, and ICU stay in the last 6 months and 30 days of life by number of serious illness diagnoses. For cancer and dementia, we examined the effect of an additional serious illness.We included a population of 1372 adults who were, on average, 84 years, 52% female, and 96% white. Approximately 41% had multiple serious illnesses. Compared to older adults with 1 serious illness diagnosis, rates of hospitalization, and ICU stay for adults with 2 or ≥3 serious illness diagnoses were at least 1.5 times higher in the last 6 months and the last 30 days of life. Rates of ED visits were significantly higher for older adults with 2 or ≥3 serious illness diagnoses in the last 6 months of life, but only higher for those with ≥3 versus 1 serious illness diagnosis in the last 30 days of life. For both cancer and dementia, rates of ED visits, hospitalization and ICU stay were lower for the condition alone than when an additional serious illness diagnosis was present.Having multiple serious illnesses increases the risk of health care utilization at the end of life.

IMPORTANCE: Despite availability of safe and effective human papillomavirus (HPV) vaccines, vaccination uptake remains low in the U.S. Research examining the impact of neighborhood socioeconomic status on HPV vaccination may help target interventions. OBJECTIVE: To examine the association between area deprivation and HPV vaccine initiation and completion. DESIGN, SETTING, PARTICIPANTS: Retrospective cohort study of individuals aged 11–18 years residing in the upper Midwest region. Receipt of HPV vaccination was examined over a three-year follow-up period (01/01/2016–12/31/2018). MAIN OUTCOMES AND MEASURES: Outcomes of interest were initiation and completion of HPV vaccination. Demographic data were collected from the Rochester Epidemiology Project (REP). Area-level socioeconomic disadvantage was measured by calculating an Area Deprivation Index (ADI) score for each person, a measure of socioeconomic disadvantage derived from American Community Survey data. Multivariable mixed effect Cox proportional hazards models were used to examine the association of ADI quartiles (Q1-Q4) with HPV vaccine series initiation and completion given initiation. RESULTS: Individuals residing in census block groups with higher deprivation had significantly lower likelihood of HPV vaccine initiation (Q2: HR=0.91, 0.84–0.99 Q3: HR= 0.83, 0.76–0.90; Q4: HR=0.84, 0.74–0.96) relative to those in the least-deprived block groups (Q1). Similarly, those living in block groups with higher deprivation had significantly lower likelihood of completion (Q2: HR=0.91, 0.86–0.97; Q3: HR= 0.87, 0.81–0.94; Q4: HR=0.82, 0.74–0.92) compared to individuals in the least-deprived block groups (Q1). CONCLUSIONS AND RELEVANCE: Lower probability of both HPV vaccine-series initiation and completion were observed in areas with greater deprivation. Our results can inform allocation of resources to increase HPV vaccination rates in our primary care practice and provide an example of leveraging public data to inform similar efforts across diverse health systems.

OBJECTIVE: To determine the association of socioeconomic status at the census block group level with chronic conditions, and to determine whether the associations differ by age, sex, race, or ethnicity. PATIENTS AND METHODS: Adults aged ≥20 years on April 1, 2015 from 7 counties in Southern Minnesota were identified using the Rochester Epidemiology Project records-linkage system. We estimated the prevalence of 19 chronic conditions (7 cardiometabolic, 7 other somatic, and 5 mental health conditions) at the individual level, and a composite measure of neighborhood socioeconomic disadvantage (the area deprivation index; ADI) at the census block group level (n=249). RESULTS: Among the 197,578 persons in our study, 46.7% were male, 49.5% were aged ≥50 years, 12.3% were of non-White race, and 5.3% were Hispanic. The risk of most chronic conditions increased with increasing ADI. For each cardiometabolic condition and most other somatic and mental health conditions, the pattern of increasing risk across ADI quintiles was attenuated, or there was no association across quintiles of ADI in the oldest age group (ages ≥70 years). Stronger associations between ADI and several cardiometabolic, other somatic, and mental health conditions were observed in women. CONCLUSION: Higher ADI was associated with increased risk of most chronic conditions, with more pronounced associations in younger persons. For some chronic conditions, the associations were stronger in women. Our findings underscore the importance of recognizing the overall and potentially differential impact of area-level deprivation on chronic disease outcomes for diverse populations.

Objective To examine associations of cumulative exposure to proton pump inhibitors (PPIs) with total cardiovascular disease (CVD; composed of stroke, coronary heart disease, and heart failure [HF]) and HF alone in a cohort study of White and African American participants of the Atherosclerosis Risk in Communities (ARIC) study. Methods Use of PPIs was assessed by pill bottle inspection at visit 1 (January 1, 1987 to 1989) and up to 10 additional times before baseline (visit 5; 2011 to 2013). We calculated cumulative exposure to PPIs as days of use from visit 1 to baseline. Participants (n=4346 free of total CVD at visit 5; mean age, 75 years) were observed for incident total CVD and HF events through December 31, 2016. We used Cox regression to measure associations of PPIs with total CVD and HF. Results After adjustment for potential confounding variables, participants with a cumulative exposure to PPIs of more than 5.1 years had a 2.02-fold higher risk of total CVD (95% CI, 1.50 to 2.72) and a 2.21-fold higher risk of HF (95% CI, 1.51 to 3.23) than nonusers. Conclusion Long-term PPI use was associated with twice the risk of total CVD and HF compared with nonusers. Our findings are in concordance with other research and suggest another reason to be cautious of PPI overuse.

Objectives To compare the agreement between percentile ranks from 4 multi-morbidity scores. Design Population-based descriptive study. Setting Olmsted County, Minnesota (USA). Participants We used the medical records-linkage system of the Rochester Epidemiology Project (REP; http://www.rochesterproject.org ) to identify all residents of Olmsted County, Minnesota who reached one or more birthdays between 1 January 2005 and 31 December 2014 (10 years). Methods For each person, we calculated 4 multi-morbidity scores using readily available diagnostic code lists from the US Department of Health and Human Services, the Clinical Classifications Software, and the Elixhauser Comorbidity Index. We calculated scores using diagnostic codes received in the 5 years before the index birthday and fit quantile regression models across age and separately by sex to transform unweighted, simple counts of conditions into percentile ranks as compared to peers of same age and of same sex. We compared the percentile ranks of the 4 multi-morbidity scores using intra-class correlation coefficients (ICCs). Results We assessed agreement in 181,553 persons who reached a total of 1,075,433 birthdays at ages 18 years through 85 years during the study period. In general, the percentile ranks of the 4 multi-morbidity scores exhibited high levels of agreement in 6 score-to-score pairwise comparisons. The agreement increased with older age for all pairwise comparisons, and ICCs were consistently greater than 0.65 at ages 50 years and older. Conclusions The assignment of percentile ranks may be a simple and intuitive way to assess the underlying trait of multi-morbidity across studies that use different measures.

Context Metformin is the first-line drug for treating diabetes but has a high failure rate. Objective To identify demographic and clinical factors available in the electronic health record (EHR) that predict metformin failure. Methods A cohort of patients with at least one abnormal diabetes screening test that initiated metformin was identified at three sites (Arizona, Mississippi, and Minnesota). We identified 22,047 metformin initiators (48% female, mean age of 57 ± 14 years) including 2141 African Americans, 440 Asians, 962 Other/Multi-racials, 1539 Hispanics, and 16,764 Non-Hispanic whites. We defined metformin failure as either the lack of a target hemoglobin A1c ( Results In this large diverse population, we observed a high rate of metformin failure (33%). The XGBoost model that included baseline hemoglobin A1c, age, sex, and race/ethnicity corresponded to high discrimination performance (C-index of 0.731; 95% CI 0.722, 0.740) for risk of metformin failure. Baseline hemoglobin A1c corresponded to the largest feature performance with higher levels associated with metformin failure. The addition of other clinical factors improved model performance (0.745; 95% CI 0.737, 0.754, p Conclusions Baseline hemoglobin A1c was the strongest predictor of metformin failure and additional factors substantially improved performance suggesting that routinely available clinical data could be used to identify patients at high risk of metformin failure who might benefit from closer monitoring and earlier treatment intensification.

ObjectiveCeramides have been associated with several ageing-related conditions but have not been studied as a general biomarker of multimorbidity (MM). Therefore, we determined whether ceramide levels are associated with the rapid development of MM.DesignRetrospective cohort study.SettingMayo Clinic Biobank.Participants1809 persons in the Mayo Clinic Biobank ≥65 years without MM at the time of enrolment, and with ceramide levels assayed from stored plasma.Primary outcome measurePersons were followed for a median of 5.7 years through their medical records to identify new diagnoses of 20 chronic conditions. The number of new conditions was divided by the person-years of follow-up to calculate the rate of accumulation of new chronic conditions.ResultsHigher levels of C18:0 and C20:0 were associated with a more rapid rate of accumulation of chronic conditions (C18:0 z score RR: 1.30, 95% CI: 1.10 to 1.53; C20:0 z score RR: 1.26, 95% CI: 1.07 to 1.49). Higher C18:0 and C20:0 levels were also associated with an increased risk of hypertension and coronary artery disease.ConclusionsC18:0 and C20:0 were associated with an increased risk of cardiometabolic conditions. When combined with biomarkers specific to other diseases of ageing, these ceramides may be a useful component of a biomarker panel for predicting accelerated ageing.

Minji K Lee,1 Jennifer L St Sauver,1,2 Roger T Anderson,3 Mark Linzer,4 David T Eton1,2 1Robert D. and Patricia E. Kern Center for the Science of Health Care Delivery, Mayo Clinic, Rochester, MN, USA; 2Department of Health Sciences Research, Mayo Clinic, Rochester, MN, USA; 3Department of Public Health Sciences, University of Virginia School of Medicine, Charlottesville, VA, USA; 4Hennepin Healthcare, Minneapolis, Minnesota, USA, University of Minnesota Medical School, Minneapolis, MN, USACorrespondence: Minji K LeeRobert D. and Patricia E. Kern Center for the Science of Health Care Delivery, Mayo Clinic, Harwick Building, Second Floor, 200 First St SW, Rochester, MN 55905, USATel +1 507 284 1310Email Lee.Minji@mayo.eduDavid T Eton Email eton.david@mayo.eduPurpose: To examine the factor structure and differential item functioning (DIF) of the Patient Experience with Treatment and Self-management (PETS version 2.0), a measure of treatment burden.Patients and Methods: Version 2.0 of the PETS has 60 items, extending the previously-validated 48-item version 1.0 by three domains (nine items) and three additional items in an existing domain. We conducted confirmatory factor analyses (CFA) on survey responses of 439 community-dwelling adults living with multiple chronic conditions who completed PETS version 2.0, using R packages, “lavaan” and “semTools.” We tested fit of second-order factors to explore simplifying the reporting of PETS scores. We examined DIF for the two second-order factors with “lordif” R package, testing groups by gender, education, and health literacy, using the McFadden pseudo R2 change criterion of ≥ 0.02 to flag items with DIF. Cronbach’s alpha and the intraclass correlation coefficient (ICC) were used to determine the reliability of PETS domains.Results: The first-order CFA model featuring 12 multi-item domains had an excellent fit (Comparative Fit Index [CFI]=0.989), as did the second-order CFA model (CFI=0.987), specifying two superordinate factors of treatment burden (workload and impact). Items in the workload and impact second-order factors did not show any DIF across gender, education, and health literacy groups as shown by McFadden pseudo R2 changes < 0.02. Cronbach’s alphas for all multi-item domain scales were ≥ 0.80, and ICCs of ten scales were ≥ 0.70, meeting the threshold for adequate test–retest reliability.Conclusion: Findings support the construct validity and reliability of PETS version 2.0. The fit of a factor model featuring superordinate (ie, second-order) factors of workload and impact supports index scoring that will simplify reporting of PETS scores. DIF analyses indicate that items from these indices can be interpreted in the same way, regardless of gender, education, or health literacy.Keywords: factor structure, patient-reported outcomes, multi-morbidity, psychometric testing, questionnaire, validation

Objective To assess the potential impact of Pharmacogenomic (PGx) variation in cytochrome P450 2D6 (CYP2D6) enzyme function, using loss in quality-adjusted life years (QALYs) associated with treatment problems, and the willingness to pay to avoid treatment problems from patients’ and payers’ perspectives. Patients and Methods The study included patients prescribed tramadol or codeine, or both, between January 1, 2005, and December 31, 2017. Demographic information and adverse drug events, including adverse drug events and poor pain control, were collected from the electronic health records using natural language processing techniques and review by trained abstractors. Patients’ willingness to pay and QALY estimates were based on comprehensive literature review. The CYP2D6 phenotypes were divided into 4 groups: ultra-rapid metabolizers, normal metabolizers, intermediate metabolizers, and poor metabolizers. Results Among the 2860 identified patients, 63 (2%) were ultrarapid metabolizers, 1449 (50%) were normal metabolizers, 1155 (40%) were intermediate metabolizers, and 193 (7%) were poor metabolizers. The patients’ average estimated willingness-to-pay value to avoid treatment problems was $23 per month; poor metabolizers developed problems with the highest estimated willingness-to-pay value ($32 per month). The mean QALY loss among all patients was 0.024 QALYs (8.8 healthy days); poor metabolizers had the highest loss (0.027 QALYs, 9.9 healthy days). Conclusion Patients with various phenotypes developed different treatment problem profiles. Poor CYP2D6 metabolizers developed problems with highest willingness to pay, and they might potentially benefit most from PGx-guided treatment and problem prevention.

Background Delirium is underdiagnosed in clinical practice and is not routinely coded for billing. Manual chart review can be used to identify the occurrence of delirium; however, it is labor-intensive and impractical for large-scale studies. Natural language processing (NLP) has the capability to process raw text in electronic health records (EHRs) and determine the meaning of the information. We developed and validated NLP algorithms to automatically identify the occurrence of delirium from EHRs. Methods This study used a randomly selected cohort from the population-based Mayo Clinic Biobank (N = 300, age ≥65). We adopted the standardized evidence-based framework confusion assessment method (CAM) to develop and evaluate NLP algorithms to identify the occurrence of delirium using clinical notes in EHRs. Two NLP algorithms were developed based on CAM criteria: one based on the original CAM (NLP-CAM; delirium vs no delirium) and another based on our modified CAM (NLP-mCAM; definite, possible, and no delirium). The sensitivity, specificity, and accuracy were used for concordance in delirium status between NLP algorithms and manual chart review as the gold standard. The prevalence of delirium cases was examined using International Classification of Diseases, 9th Revision (ICD-9), NLP-CAM, and NLP-mCAM. Results NLP-CAM demonstrated a sensitivity, specificity, and accuracy of 0.919, 1.000, and 0.967, respectively. NLP-mCAM demonstrated sensitivity, specificity, and accuracy of 0.827, 0.913, and 0.827, respectively. The prevalence analysis of delirium showed that the NLP-CAM algorithm identified 12 651 (9.4%) delirium patients, the NLP-mCAM algorithm identified 20 611 (15.3%) definite delirium cases, and 10 762 (8.0%) possible cases. Conclusions NLP algorithms based on the standardized evidence-based CAM framework demonstrated high performance in delineating delirium status in an expeditious and cost-effective manner.

PURPOSE: The purpose of this study was to test the known-groups validity and responsiveness to change of the Patient Experience with Treatment and Self-management (PETS, vs. 2.0), a measure of treatment burden. METHODS: The PETS and other standard measures were mailed at baseline and 12-month follow-up to adults living with multiple chronic conditions in southeast Minnesota (USA). A sample of 365 people (mean age = 62.1 years) completed both surveys. Baseline, 12-month, and changes in PETS burden scores were examined. Clinical anchors used to test validity included number of diagnoses (2-4 vs. 5+), mental health diagnosis (yes/no), medication adherence and health literacy (suboptimal/optimal), and changes in self-efficacy, global physical and global mental health (worsening/improving). Independent samples t-tests were used to compare scores. RESULTS: PETS scales showed good internal consistency (alphas ≥ 0.80). There were few differences across number of diagnoses, but having a mental health diagnosis was associated with higher baseline PETS burden scores (Ps < .05). Suboptimal medication adherence and health literacy over time were associated with worse 12-month PETS burden scores (Ps < .05). Compared with improvements, declines over time in self-efficacy, global physical health, and global mental health were each associated with worsening change scores on PETS impact summary, medical expenses, and bother due to medication reliance and medication side effects (Ps < .05). CONCLUSION: Among multi-morbid adults, the PETS demonstrated evidence of known-groups validity and responsiveness to change across both objective (e.g., mental health diagnoses) and subjective anchors (e.g., changes in self-efficacy, global physical and mental health).

Aim: Pharmacogenomics (PGx) holds potential to improve patient treatment; yet, effective patient educational materials are limited. Materials & methods: Using a ‘think aloud’ technique, we sought to understand comprehension and perceptions of a multimedia PGx results packet including a cover letter with QR code to an educational video, brochure and prototype report in the context of PGx case vignettes. Results: The cover letter and video components were viewed less favorably due to excess detail, complex jargon and technology challenges. Recommendations were to enhance comprehension and utility and to customize materials to each patient’s medications or disease conditions. Conclusion: Educational materials were revised to improve comprehension and usability, and diminish concerns to better prepare patients to understand the importance of discussing test results with their provider.

Purpose To examine the evidence on the cost-effectiveness of implementing pharmacogenomics (PGx) in cardiovascular disease (CVD) care. Methods We conducted a systematic review using multiple databases from inception to 2018. The titles and abstracts of cost-effectiveness studies on PGx-guided treatment in CVD care were screened, and full texts were extracted. Results We screened 909 studies and included 46 to synthesize. Acute coronary syndrome and atrial fibrillation were the predominantly studied conditions (59%). Most studies (78%) examined warfarin–CYP2C9/VKORC1 or clopidogrel–CYP2C19. A payer’s perspective was commonly used (39%) for cost calculations, and most studies (46%) were US-based. The majority (67%) of the studies found PGx testing to be cost-effective in CVD care, but cost-effectiveness varied across drugs and conditions. Two studies examined PGx panel testing, of which one examined pre-emptive testing strategies. Conclusion We found mixed evidence on the cost-effectiveness of PGx in CVD care. Supportive evidence exists for clopidogrel–CYP2C19 and warfarin–CYP2C9/VKORC1, but evidence is limited in other drug–gene combinations. Gaps persist, including unclear explanation of perspective and cost inputs, underreporting of study design elements critical to economic evaluations, and limited examination of PGx panel and pre-emptive testing for their cost-effectiveness. This review identifies the need for further research on economic evaluations of PGx implementation.

Background Several small studies have previously investigated associations between the cytochrome P450 2D6 (CYP2D6) metabolism and response to opioids. We used a large sample of patients to study associations between CYP2D6 phenotypes and estimated CYP2D6 enzymatic activity scores with pain control and adverse reactions related to codeine and tramadol use. We conducted additional analyses to determine whether our results were consistent among men and women. Methods We used data from 2,877 participants in the RIGHT Protocol who were prescribed codeine and/or tramadol between 01/01/2005 and 12/31/2017 and who were not prescribed CYP2D6 inhibitors within 1 year prior to the opioid prescription. CYP2D6 phenotype categories were condensed into four groups: (1) Ultra-rapid and Rapid (n = 61), (2) Normal and Intermediate to Normal (n = 1,448), (3) Intermediate and Intermediate to Poor (n = 1,175), and (4) Poor metabolizer status (n = 193). Opioid-related outcomes included indications of poor pain control or adverse reactions related to medication use. We modeled the risk of each outcome using logistic regression, adjusting for age, sex, race, and ethnicity. Results The results revealed a trend from poor to ultra-rapid and rapid CYP2D6 phenotypes in which the risk of adverse reactions incrementally increased and the risk of poor pain control incrementally decreased. This trend reached statistical significance among female (but not male) participants. Among normal and intermediate to normal metabolizers, a larger proportion of women experienced adverse reactions relative to men. Discussion We replicated and extended the findings of previous research indicating associations between CYP2D6 phenotypes and response to opioids. In addition, the observed associations were stronger in women than in men. We recommend sex differences to be factored in future research investigating associations between pharmacogenomics and response to medications.

Background Persons with low socioeconomic status may be disproportionately at risk for multimorbidity. Methods Adults aged ≥20 years on 4/1/2015 from 7 counties in Minnesota were identified using the Rochester Epidemiology Project (population-based sample). A composite measure of neighborhood socioeconomic disadvantage, the area deprivation index (ADI), was estimated at the census block group level (n = 251). The prevalence of 21 chronic conditions was obtained to calculate the proportion of persons with multimorbidity (≥2 chronic conditions) and severe multimorbidity (≥5 chronic conditions). Hierarchical logistic regression was used to estimate the association of ADI with multimorbidity and severe multimorbidity using odds ratios (OR). Results Among 198,941 persons (46.7% male, 30.6% aged ≥60 years), the age- and sex-standardized (to the United States 2010 census) median prevalence (Q1, Q3) was 23.4% (21.3%, 25.9%) for multimorbidity and 4.8% (4.0%, 5.7%) for severe multimorbidity. Compared with persons in the lowest quintile of ADI, persons in the highest quintile had a 50% increased risk of multimorbidity (OR 1.50, 95% CI 1.39–1.62) and a 67% increased risk of severe multimorbidity (OR 1.67, 95% CI 1.51–1.86) after adjusting for age, sex, race, and ethnicity. Associations were stronger after further adjustment for individual level of education; persons in the highest quintile had a 78% increased risk of multimorbidity (OR 1.78, 95% CI 1.62–1.96) and a 92% increased risk of severe multimorbidity (OR 1.92, 95% CI 1.72–2.13). There was evidence of interactions between ADI and age, between ADI and sex, and between ADI and education. After age 70 years, no difference in the risk of multimorbidity was observed across quintiles of ADI. The pattern of increasing multimorbidity with increasing ADI was more pronounced in women. Finally, there was less variability across quintiles of ADI for the most highly educated group. Conclusions Higher ADI was associated with increased risk of multimorbidity, and the associations were strengthened after adjustment for individual level of education, suggesting that neighborhood context plays a role in health above and beyond individual measures of socioeconomic status. Furthermore, associations were more pronounced in younger persons and women, highlighting the importance of interventions to prevent chronic conditions in younger women, in particular.

Objectives Obesity is a potentially modifiable risk factor that has been consistently associated with the development and progression of multi-morbidity (MM). However, obesity may be more problematic for some persons compared to others because of interactions with other risk factors. Therefore, we studied the effect of interactions between patient characteristics and overweight and obesity on the rate of accumulation of MM. Methods We studied 4 cohorts of persons ages 20-, 40-, 60-, and 80-years residing in Olmsted County, Minnesota between 2005 and 2014 using the Rochester Epidemiology Project (REP) medical records-linkage system. Body mass index, sex, race, ethnicity, education, and smoking status were extracted from REP indices. The rate of accumulation of MM was calculated as the number of new chronic conditions accumulated per 10 person years through 2017. Poisson rate regression models were used to identify associations between characteristics and rate of MM accumulation. Additive interactions were summarized using relative excess risk due to interaction, attributable proportion of disease, and the synergy index. Results Greater than additive synergistic associations were observed between female sex and obesity in the 20- and 40-year cohorts, between low education and obesity in the 20-year cohort (both sexes), and between smoking and obesity in the 40-year cohort (both sexes). Conclusions Interventions targeted at women, persons with lower education, and smokers who also have obesity may result in the greatest reduction in the rate of MM accumulation. However, interventions may need to focus on persons prior to mid-life to have the greatest effect.

PURPOSE: The Mayo-Baylor RIGHT 10K Study enabled preemptive, sequence-based pharmacogenomics (PGx)-driven drug prescribing practices in routine clinical care within a large cohort. We also generated the tools and resources necessary for clinical PGx implementation and identified challenges to be overcome. Furthermore, we measured the frequency of both common genetic variation for which clinical guidelines already exist and that of rare variation that could be detected by DNA sequencing, rather than genotyping. METHODS: Targeted oligonucleotide-capture sequencing of 77 “pharmacogenes” was performed using DNA from 10,077 consented Mayo Clinic Biobank volunteers. The resulting predicted drug response related phenotypes for 13 genes, including CYP2D6 and HLA, affecting 21 drug-gene pairs, were deposited preemptively in the Mayo electronic health record (EHR). RESULTS: For the 13 pharmacogenes of interest, the genomes of 79% of participants carried clinically actionable variants in 3 or more genes and DNA sequencing identified an average of 3.3 additional conservatively predicted deleterious variants that would not have been evident using genotyping. CONCLUSIONS: Implementation of preemptive rather than reactive, sequence-based rather than genotype-based PGx prescribing revealed nearly universal patient applicability and required integrated institution-wide resources to fully realize individualized drug therapy and to demonstrate more efficient use of health care resources.

Objective To identify risk factors associated with severe COVID-19 infection in a defined Midwestern US population overall and within different age groups. Patients and Methods We used the Rochester Epidemiology Project research infrastructure to identify persons residing in a defined 27-county Midwestern region who had positive results on polymerase chain reaction tests for COVID-19 between March 1, 2020, and September 30, 2020 (N=9928). Age, sex, race, ethnicity, body mass index, smoking status, and 44 chronic disease categories were considered as possible risk factors for severe infection. Severe infection was defined as hospitalization or death caused by COVID-19. Associations between risk factors and severe infection were estimated using Cox proportional hazard models overall and within 3 age groups (0 to 44, 45 to 64, and 65+ years). Results Overall, 474 (4.8%) persons developed severe COVID-19 infection. Older age, male sex, non-White race, Hispanic ethnicity, obesity, and a higher number of chronic conditions were associated with increased risk of severe infection. After adjustment, 36 chronic disease categories were significantly associated with severe infection. The risk of severe infection varied significantly across age groups. In particular, persons 0 to 44 years of age with cancer, chronic neurologic disorders, hematologic disorders, ischemic heart disease, and other endocrine disorders had a greater than 3-fold increased risk of severe infection compared with persons of the same age without those conditions. Associations were attenuated in older age groups. Conclusion Older persons are more likely to experience severe infections; however, severe cases occur in younger persons as well. Our data provide insight regarding younger persons at especially high risk of severe COVID-19 infection.

PurposeThe depth and breadth of clinical data within electronic health record (EHR) systems paired with innovative machine learning methods can be leveraged to identify novel risk factors for complex diseases. However, analysing the EHR is challenging due to complexity and quality of the data. Therefore, we developed large electronic population-based cohorts with comprehensive harmonised and processed EHR data.ParticipantsAll individuals 30 years of age or older who resided in Olmsted County, Minnesota on 1 January 2006 were identified for the discovery cohort. Algorithms to define a variety of patient characteristics were developed and validated, thus building a comprehensive risk profile for each patient. Patients are followed for incident diseases and ageing-related outcomes. Using the same methods, an independent validation cohort was assembled by identifying all individuals 30 years of age or older who resided in the largely rural 26-county area of southern Minnesota and western Wisconsin on 1 January 2013.Findings to dateFor the discovery cohort, 76 255 individuals (median age 49; 53% women) were identified from which a total of 9 644 221 laboratory results; 9 513 840 diagnosis codes; 10 924 291 procedure codes; 1 277 231 outpatient drug prescriptions; 966 136 heart rate measurements and 1 159 836 blood pressure (BP) measurements were retrieved during the baseline time period. The most prevalent conditions in this cohort were hyperlipidaemia, hypertension and arthritis. For the validation cohort, 333 460 individuals (median age 54; 52% women) were identified and to date, a total of 19 926 750 diagnosis codes, 10 527 444 heart rate measurements and 7 356 344 BP measurements were retrieved during baseline.Future plansUsing advanced machine learning approaches, these electronic cohorts will be used to identify novel sex-specific risk factors for complex diseases. These approaches will allow us to address several challenges with the use of EHR.

Background: In 2012, updated cervical cancer screening recommendations were released with consensus on Papanicolaou (Pap) testing every 3 years for women age 21–65 years or Pap–human papillomavirus (HPV) cotesting at 5-year intervals for women age 30–65 years. Primary study aims: Assess current use of Pap-HPV cotesting and describe local population trends over time in Pap and Pap-HPV cotesting. Secondary aim: Assess sociodemographic factors correlating with screening. Methods: We assessed Rochester Epidemiology Project data for Pap and Pap-HPV cotesting among women age 16 years and older living in Olmsted County, Minnesota, yearly from 2005 (study population n = 47,203) through 2016 (study population n = 49,510). We calculated 3-year (Pap) and 5-year (Pap-HPV) moving prevalence rates of screening as proportion of eligible population. Multivariable logistic regression was used to assess factors potentially associated with screening. Results: In 2016, 64.6% of 27,418 eligible 30- to 65-year-old women were up to date with cervical cancer screening; 60.8% had received Pap-HPV cotest screening. Significant declines in Pap completion rates over time were observed in all age groups, including an unexpected decline in 21- to 29-year-old women. Coincident with decreasing Pap screening rates, Pap-HPV cotesting significantly increased among women age 30–65 years, from 10.0% in 2007 to 60.8% in 2016. Conclusions: This suggests increasing adoption of 2012 screening recommendations in the 30- to 65-year-old population. However, decline in Pap screening among 21- to 29-year-old women is concerning. Disparities by race, ethnicity, smoking status, and comorbidity level were observed. Results suggest need for multilevel patient and clinician interventions to increase cervical cancer screening adherence.

Background: Prevention strategies for Alzheimer’s disease and Alzheimer’s disease related dementias (AD/ADRDs) are urgently needed for reducing incidence. Intra-patient variability in lipid levels is a potentially modifiable risk factor for incident AD/ADRD. Although regular lipid measurements are a part of common clinical practice and longitudinal data routinely available in electronic health records (EHR), research examining this association between AD/ADRD and lipid variability across multiple lipid types remains scarce. Methods: All residents living in Olmsted County, MN on 1/1/2006 age ≥60 years without an AD/ADRD diagnosis were identified using Centers for Medicare & Medicaid Services diagnostic codes. Persons with ≥3 lipid measurements (total cholesterol or triglycerides) in the 5 years prior to index date were retained. Lipid variability was quantified using variability independent of the mean (VIM). Models were adjusted for traditional risk factors. Associations between lipid variability quintiles and incident AD/ADRD were assessed using Cox proportional hazards regression. Multiple imputation was used for missing covariates. Participants were followed through 2018 for incident AD/ADRD. Results: The final analysis included data on 11,551 participants with total cholesterol and 11,502 participants with triglycerides. Participants had a mean age of 71 (range 60-99) years, and were primarily white (96%). Females (54%) were also slightly more frequent than males. Median follow up was 12.9 years (range: 0-13). Participants in the highest quintile of variability for total cholesterol and triglycerides had a 20% increased risk of incident AD/ADRD. Similar results were found in the subset with complete covariate data. Conclusion: In a large EHR derived cohort, persons in the highest quintile of lipid variation had an increased risk of incident AD/ADRD. Further studies to identify the mechanisms behind this risk factor and replication of these results across a more diverse population are needed.

Background: Larger within-patient variability of lipid levels has been associated with an increased risk of cardiovascular disease (CVD). However, measures of lipid variability are not currently used clinically. We investigated the feasibility of calculating lipid variability within a large electronic health record (EHR)-based population cohort and assessed associations with incident CVD. Methods: We identified all individuals ≥40 years of age who resided in Olmsted County, MN on 1/1/2006 (index date) without prior CVD. CVD was defined as myocardial infarction, coronary artery bypass graft surgery, percutaneous coronary intervention or stroke. Patients with ≥3 measurements of total cholesterol, low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), and/or triglycerides during the 5 years before the index date were retained in the analyses. Lipid variability was calculated using variability independent of the mean (VIM). Patients were followed through 9/30/2017 for incident CVD (including CVD death). Cox regression was used to investigate the association between quintiles of lipid VIMs and incident CVD. Results: We identified 18,642 individuals (mean age 60; 55% female) who were free of CVD at baseline and VIM calculated for at least one lipid measurement. After adjustment, those in the highest VIM quintiles of total cholesterol had a 25% increased risk of CVD (Q5 vs. Q1 HR: 1.25, 95% CI: 1.08-1.45; Table). We observed similar results for LDL-C (Q5 vs. Q1 HR: 1.20, 95% CI: 1.04-1.39) and HDL-C (Q5 vs. Q1 HR: 1.25, 95% CI: 1.09-1.43). There was no association between triglyceride variability quintiles and CVD risk. Conclusion: In a large EHR-based population cohort, high variability in total cholesterol, HDL-C and LDL-C was associated with an increased risk of CVD, independently of traditional risk factors, suggesting it may be a target for intervention. Lipid variability can be calculated in the EHR environment but more research is needed to determine its clinical utility.

Objective To assess the validity of the US Department of Health and Human Services (DHHS) definition of multimorbidity using International Classification of Diseases, ninth edition (ICD-9) codes from administrative data. Design Cross-sectional comparison of two ICD-9 billing code algorithms to data abstracted from medical records. Setting Olmsted County, Minnesota, USA. Participants An age-stratified and sex-stratified random sample of 1509 persons ages 40–84 years old residing in Olmsted County on 31 December 2010. Study measures Seventeen chronic conditions identified by the US DHHS as important in studies of multimorbidity were identified through medical record review of each participant between 2006 and 2010. ICD-9 administrative billing codes corresponding to the 17 conditions were extracted using the Rochester Epidemiology Project records-linkage system. Persons were classified as having each condition using two algorithms: at least one code or at least two codes separated by more than 30 days. We compared the ICD-9 code algorithms with the diagnoses obtained through medical record review to identify persons with multimorbidity (defined as ≥2, ≥3 or ≥4 chronic conditions). Results Use of a single code to define each of the 17 chronic conditions resulted in sensitivity and positive predictive values (PPV) ≥70%, and in specificity and negative predictive values (NPV) ≥70% for identifying multimorbidity in the overall study population. PPV and sensitivity were highest in persons 65–84 years of age, whereas NPV and specificity were highest in persons 40–64 years. The results varied by condition, and by age and sex. The use of at least two codes reduced sensitivity, but increased specificity. Conclusions The use of a single code to identify each of the 17 chronic conditions may be a simple and valid method to identify persons who meet the DHHS definition of multimorbidity in populations with similar demographic, socioeconomic, and health care characteristics.

The study of sex-specific genetic associations with opioid response may improve the understanding of inter-individual variability in pain treatments. We investigated sex-specific associations between genetic variation and opioid response. We identified participants in the RIGHT Study prescribed codeine, tramadol, hydrocodone, and oxycodone between 01/01/2005 and 12/31/2017. Prescriptions were collapsed into codeine/tramadol and hydrocodone/oxycodone. Outcomes included poor pain control and adverse reactions within six weeks after prescription date. We performed gene-level and single-variant association analyses stratified by sex. We included 7169 non-Hispanic white participants and a total of 1940 common and low-frequency variants (MAF 0.01). Common variants in MACROD2 (rs76026520), CYP1B1 (rs1056837, rs1056836), and CYP2D6 (rs35742686) were associated with outcomes. At the gene level, FAAH, SCN1A, and TYMS had associations for men and women, and NAT2, CYP3A4, CYP1A2, and SLC22A2 had associations for men only. Our findings highlight the importance of considering sex in association studies on opioid response.