Characterizing Bias Due to Differential Exposure Ascertainment in Electronic Health Record Data.

Data derived from electronic health records (EHR) are heterogeneous with availability of specific measures dependent on the type and timing of patients' healthcare interactions. This creates a challenge for research using EHR-derived exposures because gold-standard exposure data, determined by a definitive assessment, may only be available for a subset of the population. Alternative approaches to exposure ascertainment in this case include restricting the analytic sample to only those patients with gold-standard exposure data available (exclusion); using gold-standard data, when available, and using a proxy exposure measure when the gold standard is unavailable (best available); or using a proxy exposure measure for everyone (common data). Exclusion may induce selection bias in outcome/exposure association estimates, while incorporating information from a proxy exposure via either the best available or common data approaches may result in information bias due to measurement error. The objective of this paper was to explore the bias and efficiency of these three analytic approaches across a broad range of scenarios motivated by a study of the association between chronic hyperglycemia and five-year mortality in an EHR-derived cohort of colon cancer survivors. We found that the best available approach tended to mitigate inefficiency and selection bias resulting from exclusion while suffering from less information bias than the common data approach. However, bias in all three approaches can be severe, particularly when both selection bias and information bias are present. When risk of either of these biases is judged to be more than moderate, EHR-based analyses may lead to erroneous conclusions.

Cancer disparities between mainland and island Puerto Ricans.

OBJECTIVES: Examination of cancer rates in a single Hispanic subgroup-Puerto Ricans- and comparison of incidence rates among mainland Puerto Ricans living in the United States, island Puerto Ricans in Puerto Rico, and U.S. non-Hispanic whites to reveal ethnic-specific cancer patterns and disparities in Puerto Ricans. METHODS: Incidence data were obtained from the cancer registries of Puerto Rico and three U.S. northeastern states (New York, New Jersey, and Connecticut) with a high density of mainland Puerto Ricans. Age-adjusted rates were compared by standardized rate ratios (SRRs). RESULTS: Total cancer incidence was the lowest in island Puerto Ricans, intermediate for mainland Puerto Ricans, and highest in U.S. non-Hispanic whites. Compared to mainland Puerto Ricans, islanders had significantly lower rates (p<0.05) for major cancers-lung (SRRs=0.36 in males and 0.29 in females), prostate (SRR=0.71), female breast (SRR=0.73), and colon-rectum (SRRs=0.74 in males and 0.65 in females)-as well as several less common cancers (urinary bladder; non-Hodgkin lymphoma; liver; kidney and renal pelvis; pancreas; thyroid; leukemia; and skin melanoma). Overall cancer rates in mainland Puerto Ricans were modestly lower than those in U.S. non-Hispanic whites, but mainland Puerto Ricans had the highest rates for stomach, liver, and cervical cancers among the three populations. CONCLUSION: Despite socioeconomic disadvantages, island Puerto Ricans have relatively low cancer incidence. Identifying contributing factors would be informative for cancer research, and understanding the reasons for increased cancer risk in their mainland counterparts would facilitate the development of ethnic-specific intervention programs.

Cancer disparities between mainland and island Puerto Ricans / Disparidades en la incidencia de cáncer en puertorriqueños residentes en los Estados Unidos y Puerto Rico

OBJECTIVES: Examination of cancer rates in a single Hispanic subgroup-Puerto Ricans-and comparison of incidence rates among mainland Puerto Ricans living in the United States, island Puerto Ricans in Puerto Rico, and U.S. non-Hispanic whites to reveal ethnic-specific cancer patterns and disparities in Puerto Ricans. METHODS: Incidence data were obtained from the cancer registries of Puerto Rico and three U.S. northeastern states (New York, New Jersey, and Connecticut) with a high density of mainland Puerto Ricans. Age-adjusted rates were compared by standardized rate ratios (SRRs). RESULTS: Total cancer incidence was the lowest in island Puerto Ricans, intermediate for mainland Puerto Ricans, and highest in U.S. non-Hispanic whites. Compared to mainland Puerto Ricans, islanders had significantly lower rates (p < 0.05) for major cancers-lung (SRRs = 0.36 in males and 0.29 in females), prostate (SRR = 0.71), female breast (SRR = 0.73), and colon-rectum (SRRs = 0.74 in males and 0.65 in females)-as well as several less common cancers (urinary bladder; non-Hodgkin lymphoma; liver; kidney and renal pelvis; pancreas; thyroid; leukemia; and skin melanoma). Overall cancer rates in mainland Puerto Ricans were modestly lower than those in U.S. non-Hispanic whites, but mainland Puerto Ricans had the highest rates for stomach, liver, and cervical cancers among the three populations. CONCLUSION: Despite socioeconomic disadvantages, island Puerto Ricans have relatively low cancer incidence. Identifying contributing factors would be informative for cancer research, and understanding the reasons for increased cancer risk in their mainland counterparts would facilitate the development of ethnic-specific intervention programs.

OBJETIVOS: Se analizaron las tasas de cáncer en un subgrupo de hispanos residentes en los Estados Unidos de América -los puertorriqueños (PRREUA) y se compararon sus tasas de incidencia con las de los puertorriqueños que residen en Puerto Rico (PRRPR) y la población estadounidense blanca sin ascendencia hispana (EUBNH) a fin de encontrar patrones de cáncer y disparidades de orden étnico específicos para los puertorriqueños. MÉTODOS: Se obtuvieron los datos de incidencia de los registros de cáncer de Puerto Rico y tres estados del nordeste de los Estados Unidos (New York, New Jersey y Connecticut) que tienen una elevada densidad de PRREUA. Se compararon las tasas ajustadas por la edad mediante las razones de las tasas estandarizadas (SRR). RESULTADOS: La incidencia total de cáncer fue menor en los PRRPR, intermedia en los PRREUA y mayor en los EUBNH. Los PRRPR presentaron tasas significativamente menores que los PRREUA (P < 0,05) en los principales tipos de cáncer -de pulmón (SRR = 0,36 en hombres; SRR = 0,29 en mujeres), próstata (SRR = 0,71), mama (SRR = 0,73) y colorrectal (SRR = 0,74 en hombres y SRR = 0,65 en mujeres)- así como en algunos tipos de cáncer menos frecuentes (de vejiga, hígado, riñón y pelvis renal, páncreas, tiroides, linfomas no Hodgkin, leucemia y melanoma de piel). En general, las tasas de cáncer en los PRREUA fueron ligeramente menores que las de los EUBNH, aunque de las tres poblaciones los PRREUA tuvieron las mayores tasas de cáncer de estómago, hígado y cervicouterino. CONCLUSIONES: A pesar de las desventajas socioeconómicas, los PRRPR tienen una menor incidencia relativa de cáncer. La identificación de los factores que contribuyen a ello podría ayudar en las investigaciones sobre cáncer, y comprender las razones del mayor riesgo de cáncer en los PRREUA podría facilitar el desarrollo de programas de intervención específicos para esta población.

Health disparities between island and mainland Puerto Ricans.

OBJECTIVE: To detect health disparities among three populations--Puerto Ricans living in Puerto Rico as well as Puerto Ricans and non-Hispanic whites living on the United States (U.S.) mainland. METHODS: Data from two similarly designed surveys conducted in 1999-2000 were analyzed. The Behavioral Risk Factor Surveillance System (BRFSS) provided data on Puerto Ricans living on the island and on non-Hispanic whites in the U.S. Another survey of Puerto Ricans living in New York City provided data on mainland Puerto Ricans. The age- and sex-standardized weighted prevalences of various health parameters (e.g., obesity, diabetes, smoking, and physical illness) and indicators of access to health care (e.g., frequencies of routine checkups and diabetes care) were compared between populations by means of standardized rate ratios (SRR). RESULTS: Puerto Ricans living on the mainland and those living on the island had a similar prevalence of obesity (21% to 22%). Compared with islanders, mainland Puerto Ricans had a higher prevalence of diabetes (SRR = 1.4; 95% confidence interval [95% CI] = 1.01 to 2.0); those with diabetes also showed higher prevalences of smoking (SRR = 4.2; 95% CI = 2.3 to 7.7) and physical illness (SRR = 1.5; 95% CI = 1.1 to 2.0) than Puerto Ricans living on the island. While mainland Puerto Ricans were similar to non-Hispanic whites in terms of their utilization of primary prevention and diabetes care, island Puerto Ricans trailed behind significantly. CONCLUSIONS: Puerto Ricans living on the U.S. mainland and those living in Puerto Rico both need to target lowering their prevalence of obesity and diabetes. For island Puerto Ricans, improved education about the significance of primary prevention and diabetes care is needed. For mainland Puerto Ricans, the accessibility of the primary health care system renders it a potentially effective venue for interventions, particularly for smoking cessation. More studies are warranted to identify factors associated with the poor health status observed in mainland Puerto Ricans.

Health disparities between island and mainland Puerto Ricans / Disparidades de salud entre puertorriqueños que viven en la isla de Puerto Rico y en tierra firme estadounidense

Objective. To detect health disparities among three populations­Puerto Ricans living in Puerto Rico as well as Puerto Ricans and non-Hispanic whites living on the United States (U.S.) mainland. Methods. Data from two similarly designed surveys conducted in 1999­2000 were analyzed. The Behavioral Risk Factor Surveillance System (BRFSS) provided data on Puerto Ricans living on the island and on non-Hispanic whites in the U.S. Another survey of Puerto Ricans living in New York City provided data on mainland Puerto Ricans. The age- and sexstandardized weighted prevalences of various health parameters (e.g., obesity, diabetes, smoking, and physical illness) and indicators of access to health care (e.g., frequencies of routine checkups and diabetes care) were compared between populations by means of standardized rate ratios (SRR). Results. Puerto Ricans living on the mainland and those living on the island had a similar prevalence of obesity (21% to 22%). Compared with islanders, mainland Puerto Ricans had a higher prevalence of diabetes (SRR = 1.4; 95% confidence interval [95% CI] = 1.01 to 2.0); those with diabetes also showed higher prevalences of smoking (SRR = 4.2; 95% CI = 2.3 to 7.7) and physical illness (SRR = 1.5; 95% CI = 1.1 to 2.0) than Puerto Ricans living on the island. While mainland Puerto Ricans were similar to non-Hispanic whites in terms of their utilization of primary prevention and diabetes care, island Puerto Ricans trailed behind significantly. Conclusions. Puerto Ricans living on the U.S. mainland and those living in Puerto Rico both need to target lowering their prevalence of obesity and diabetes. For island Puerto Ricans, improved education about the significance of primary prevention and diabetes care is needed. For mainland Puerto Ricans, the accessibility of the primary health care system renders it a potentially effective venue for interventions, particularly for smoking cessation. More studies are warranted to identify factors associated with the poor health status observed in mainland Puerto Ricans.

OBJETIVO: Detectar disparidades de salud entre tres poblaciones: puertorriqueños que viven en Puerto Rico, así como puertorriqueños y personas no hispanas de raza blanca que viven en tierra firme estadounidense. MÉTODOS: Se analizaron los datos obtenidos mediante dos encuestas de similar diseño que se realizaron en 1999–2000. El Sistema de Vigilancia de Factores de Riesgo Conductuales proporcionó datos acerca de los puertorriqueños radicados en la isla y de residentes de Estados Unidos de raza blanca que no son hispanos. Otra encuesta de puertorriqueños radicados en la Ciudad de Nueva York aportó datos acerca de los puertorriqueños que residían en tierra firme estadounidense. Se usaron las razones de las tasas estandarizadas (standardized rate ratios, SRR) para hacer las comparaciones interpoblacionales de las prevalencias ponderadas, estandarizadas por edad y sexo, de varios parámetros (obesidad, diabetes, tabaquismo y dolencias físicas) y de indicadores de acceso a la atención sanitaria (frecuencia de los exámenes de rutina y de la atención de la diabetes). RESULTADOS: Los puertorriqueños que vivían en tierra firme estadounidense y los que vivían en la isla tuvieron una prevalencia de obesidad parecida (21% a 22%). Comparados con los habitantes de la isla, los puertorriqueños radicados en tierra firme tuvieron una prevalencia de diabetes más alta (SRR = 1,4; intervalo de confianza de 95% [IC95%]: 1,01 a 2,0); los que tenían diabetes también mostraron una mayor prevalencia de tabaquismo (SRR = 4,2; IC 95%: 2,3 a 7,7) y de dolencias físicas (SRR = 1,5%; IC95%: 1,1 a 2,0) que los puertorriqueños que vivían en la isla. Mientras que los puertorriqueños en tierra firme se asemejaron a los blancos que no eran hispanos en cuanto a la utilización de servicios de prevención primaria y de atención de la diabetes, los puertorriqueños en la isla tenían cifras de utilización mucho más bajas. Conclusiones. Tanto los puertorriqueños radicados en tierra firme estadounidense como los que viven en Puerto Rico tienen que ponerse como meta reducir su prevalencia de obesidad y diabetes. En el caso de los puertorriqueños en la isla, hace falta una mejor educación en torno a la importancia de la prevención primaria y de la atención de la diabetes. En el caso de los puertorriqueños en tierra firme, la accesibilidad del sistema de atención primaria hace que este sea idóneo para llevar a cabo diversas intervenciones, particularmente contra el hábito de fumar. Se necesitan más estudios para determinar qué factores se asocian con el mal estado de salud observado en puertorriqueños radicados en tierra firme.

The interval between menarche and age of first sexual intercourse as a risk factor for subsequent HPV infection in adolescent and young adult women.

We sought to determine whether the interval between menarche and age of first sexual intercourse is associated with subsequent human papilloma virus (HPV) infection and if so, whether the association is independent of the age of first sexual intercourse. Female university students completed a survey and were screened for cervicovaginal HPV infection. HPV-positive subjects were matched to HPV-negative subjects (n = 504). Mean subject age was 20.4 +/- 2.1 years, age of first sexual intercourse 16.7 +/- 1.8 years, and interval 4.4 +/- 2.0 years. The interval was associated with HPV infection, but the association became nonsignificant in univariate stratified analyses and multivariate models estimating the association between the interval and HPV infection. A short interval is associated with HPV infection, but the association is not independent of age of first sexual intercourse. Age of first sexual intercourse should be considered an important and identifiable risk factor for subsequent HPV infection in research and clinical settings.