- Email: [email protected]
Addition of an active surveillance component to influenza-like illness surveillance in Pennsylvania, 2003–2004
AEP Vol. 14, No. 8 September 2004: 592–624
ABSTRACTS (ACE)
present racial and ethnic data. The objective of this study was to assess the extent to which recent epidemiologic research is achieving the goals of these policies. METHODS: Both federal policies and regulations and biomedical journal editor manuscript requirements governing the use of race and ethnicity data were identified. Epidemiologic studies published between December 2002 and December 2003 in the Annals of Epidemiology, American Journal of Epidemiology, and Epidemiology were reviewed for consistency with these policies and regulations. Articles were included in the review if they were human studies with individual-level outcomes and were conducted in the United States. Publications were examined to determine funding sources, rationale and methods for collecting racial and ethnic data, use of race or ethnicity in the study, data analysis methods (including use of race and ethnicity in multivariate analyses), and the extent to which race and ethnicity were addressed in presentations of results, interpretation, and discussion. RESULTS: Relevant inclusion policies and regulations were identified from U.S. Agency for Healthcare Research and Quality (AHRQ), Agency for Toxic Substances and Disease Registry/ Centers for Disease Control (ATSDR/CDC), Food and Drug Administration (FDA), Health and Human Services (HHS), and National Institutes of Health (NIH). More than 80% of studies reviewed were supported at least in part by NIH or CDC, and race and ethnicity were mentioned in more than 90% of studies; however, fewer than half of these studies provided a rationale for including race and ethnicity data or the methods used for collecting such information. Three out of four studies included race and ethnicity in statistical analyses, yet the results of these analyses were discussed in only half of studies reviewed. CONCLUSION: Although reporting of race and ethnicity in the epidemiologic literature has increased in recent years, this review finds the treatment of race and ethnicity to be incomplete and inconsistent. The use of race and ethnicity in epidemiologic studies can be improved by implementing existing policy. doi:10.1016/j.annepidem.2004.07.047
P048S ASSOCIATION BETWEEN STATE MANDATE OF CONTRACEPTIVE INSURANCE COVERAGE AND INFANT MORTALITY IN THE UNITED STATES, 2001 EL Ding, TA LaVeist, Department of Health Policy and Management, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD PURPOSE: Previous research has demonstrated that more liberal abortion policies are associated with fewer unintended births and lower neonatal mortality rates. Studies also report a link between unintended pregnancies and adverse birth outcomes such as low birth weight, a strong predictor of infant mortality. This study seeks to examine the effect of state legislative policies mandating health insurance coverage for contraceptives and its association with infant mortality. METHODS: National Center for Health Statistics data of 2001 neonatal mortality rates (NMR), infant mortality rates (IMR), and percent births of low birth weight (LBW) were compiled for U.S. states. State policies mandating insurance coverage for contraceptives were ranked as ‘‘none, limited, or comprehensive’’ by the Kaiser Family Foundation. Analyses regressed ‘‘any-coverage-
609
mandate’’ with infant health indicators, adjusting for potential confounders. Stratification for race adjusted for race-respective covariates. RESULTS: In 2001, 26 states possessed a mandate for contraceptive coverage. In crude analyses, any-coverage-mandate was associated with overall lower NMR (b Z ÿ0.63, P Z 0.04) and lower IMR (b Z ÿ1.25, P Z 0.003). In multivariate analyses, though negative regression coefficients for LBW and NMR were not significant, marginally lower IMR (b Z ÿ0.73, P Z 0.06) was found among states with any-coverage-mandates. When stratified by race, no associations were found among whites between coverage-mandate and LBW, NMR, and IMR. Among blacks, however, coverage-mandate was significantly associated with lower LBW percentage (b Z ÿ0.78, P Z 0.04), lower NMR (b Z ÿ2.78, P Z 0.01), and lower IMR (b Z ÿ3.76, P Z 0.02). CONCLUSION: State policies mandating contraceptive insurance coverage appear correlated with fewer adverse birth outcomes, especially among blacks. This corroborates the hypothesis that state policy mandating insurance coverage for contraceptives may help prevent unintended births and lower subsequent infant mortality rates by facilitating greater access of contraceptives to women of lower SES. doi:10.1016/j.annepidem.2004.07.048
P049 ADDITION OF AN ACTIVE SURVEILLANCE COMPONENT TO INFLUENZA-LIKE ILLNESS SURVEILLANCE IN PENNSYLVANIA, 2003–2004 JS Samkoff, M Brookes, Division of Infectious Disease Epidemiology and Division of Immunization, Pennsylvania Department of Health PURPOSE: The Pennsylvania Department of Health (PADOH) has participated in the CDC Influenza Surveillance Provider Network (ISPN) since 1997. In 2003, we added an active surveillance component in an effort to improve the consistency with which providers submit weekly reports of influenza-like illness (ILI) seen in their practices. Because active surveillance entailed nontrivial costs for PADOH, some measurement of its benefits was necessary to facilitate policy decisions regarding use of active ILI surveillance in future influenza seasons. METHODS: Each Tuesday from November 3, 2003, through March 2, 2004 (CDC weeks 200345 through 200409), a list of ISPN providers who did not submit an ILI report for the previous week was generated from the CDC ISPN website. Immunization nurses in the local health districts phoned these providers and offered assistance with ILI data entry. Primary measure: the difference in the percentage of ISPN providers submitting reports in 50% or more of CDC-defined influenza surveillance weeks in the 2002–2003 influenza surveillance season (no active surveillance) and the 2003–2004 season (with active surveillance). Subsidiary measure: change in reporting consistency for individual providers who participated in ISPN during both seasons, as measured by the mean and median change in % of surveillance weeks for which providers submitted reports. RESULTS: Forty-two sentinel providers participated in the entire 2003–2004 surveillance season and 66 in 2002–2003. Active surveillance resulted in an absolute increase of 46% (from 23% to 69%) in proportion of ISPN providers reporting in 50% or more of
610
ABSTRACTS (ACE)
AEP Vol. 14, No. 8 September 2004 : 592–624
surveillance weeks. Among the 37 providers who participated during both seasons, reporting consistency improved a mean of 31% (95% CI Z 20%, 43%), median 30%. CONCLUSION: Active surveillance was a valuable addition to Pennsylvania’s ISPN. Consideration of the improvement achieved versus cost incurred will be incorporated into decision making regarding future use of this active surveillance technique by PADOH. doi:10.1016/j.annepidem.2004.07.049
P050S ESTABLISHING SURGICAL POLICIES IN EAST AFRICA: RESULTS OF THE EASI-DELPHI PROJECT Massey Beveridge, Kirsteen Burton, Office of International Surgery, University of Toronto, Ontario, Canada PURPOSE: In East Africa, demand for surgical services is high: 400 surgeons provide care to a population of more than 200 million people. Given the paucity of official surgical policies in the region and the need for systematic regional planning to improve the delivery of surgical services in East Africa, as a start, basic surgical priorities needed to be identified. METHODS: Thirty-one members of the Association of Surgeons of East Africa (ASEA) participated in a Delphi process. The first stage consisted of a survey of surgical issues in East Africa. The results of the survey were circulated to the participants who were then asked to generate statements in response to the question, ‘‘What actions will most reduce the burden of surgical disease in East Africa by 2010?’’ Seventy-nine statements were received; after combining similar statements, 60 were returned to the group, who then scored the desirability and feasibility of each statement. Lower-scoring items were discarded and the remaining 25 statements were returned to the participants for ranking. The 10 statements with the highest mean scores and least variance were identified. RESULTS: Five of the top 10 priorities identified were to (i) improve opportunities for continuing medical education for practicing surgeons, (ii) introduce more surgical skills workshops for medical students and clinical officers, (iii) involve COSECSA in surgical training as well as curriculum development and certification of surgeons, (iv) provide a feedback system by which medical students and surgical trainees may evaluate their teachers, and (v) recruit and train more nurses and anesthetists. CONCLUSION: If adopted and implemented, these priorities may help the ASEA, African Ministries of Health, surgical educators, hospital administrators and individual surgeons to reduce the burden of surgical disease in East Africa by the year 2010. doi:10.1016/j.annepidem.2004.07.050
HEALTH SERVICES P051S WOMEN AND WILLINGNESS TO PARTICIPATE IN CLINICAL TRIALS: RESULTS FROM A HYPOTHETICAL RANDOMIZED CONTROL TRIAL EL Ding, NR Powe, JB Braunstein, Johns Hopkins Medical Institutions, Baltimore, MD
PURPOSE: The U.S. Congress, the Institute of Medicine, and the National Institutes of Health have all expressed great concern about insufficient representation of women in medical research, particularly within large-scale cohort and clinical safety and efficacy trials. Despite federal mandates requiring adequate representation and enrollment of women in such healthcare research, gender-related participation disparities often persist. We asked whether underrepresentation of women in clinical trials could be due to a lesser willingness to participate (WTP) on the part of women relative to men. METHODS: We randomly approached patients from 13 Maryland outpatient clinics to self-complete a survey of their WTP in a hypothetical cardiovascular chemoprevention randomized control trial (RCT). After participants rated their WTP on a five-point Likert scale (CWTP Z very likely, likely), impact of gender on WTP was analyzed from logistic regressions, adjusted for potential sociodemographic and clinical confounders. Interactions between gender and covariates were also evaluated. RESULTS: Of 1132 eligible individuals, 70% responded. By gender, 31% of women were WTP in the hypothetical RCT, compared with 38% of men. In crude analysis, women were marginally less willing to participate in the trial (odds ratio OR Z 0.74, 95% confidence interval CI Z 0.55–1.00, P Z 0.049). In the adjusted model, women indicated a significantly lower WTP than men (OR Z 0.61, 95% CI Z 0.49–0.77, P ! 0.001). Covariates interacting with gender associated with increased female WTP included lower education (!12 years education; P Z 0.001), having a sick family member (P Z 0.03), being current smoker (P Z 0.08), Caucasian race (P Z 0.10), having greater income (O$30,000/yr; P Z 0.09), and rating religion as being less important in their lives (P Z 0.04). CONCLUSION: Women appear less willing to participate in randomized control trials than men. This gender disparity in trial participation may be an obstacle to federal mandates requiring adequate representation of women in cohorts and clinical trials. doi:10.1016/j.annepidem.2004.07.051
P052 DIRECT MEDICAL COSTS OF PRETERM BIRTH FROM BIRTH TO AGE 7 YEARS: A POPULATION-BASED STUDY IN MANITOBA, CANADA CV Newburn-Cook, M Heaman, D Schopflocher, E Forget, P Jacobs, O Casiro, J Blanchard, R Sauve, LL Roos, Universities of Alberta and Manitoba, Canada PURPOSE: (i) To determine direct medical costs by gestational age and survival status from birth to age 7 years in two cohorts (16,079 newborns in 1987 and 15,853 newborns in 1993) in Manitoba, Canada, and (ii) to determine the incremental costs associated with preterm birth. The 1993 cohort was chosen to examine the impact of surfactant on survival and costs. METHODS: A longitudinal population-based study using administrative data to calculate actual costs of hospital inpatient care (based on Refined Diagnostically Related Groups), physician and lab or diagnostic costs associated with hospital stays, and outpatient care, using 1995 constant dollars. RESULTS: Costs were inversely associated with gestational age. Substantial costs were associated with the initial hospital stay and first year of life; costs declined dramatically over the next 6 years.
ABSTRACTS (ACE)
present racial and ethnic data. The objective of this study was to assess the extent to which recent epidemiologic research is achieving the goals of these policies. METHODS: Both federal policies and regulations and biomedical journal editor manuscript requirements governing the use of race and ethnicity data were identified. Epidemiologic studies published between December 2002 and December 2003 in the Annals of Epidemiology, American Journal of Epidemiology, and Epidemiology were reviewed for consistency with these policies and regulations. Articles were included in the review if they were human studies with individual-level outcomes and were conducted in the United States. Publications were examined to determine funding sources, rationale and methods for collecting racial and ethnic data, use of race or ethnicity in the study, data analysis methods (including use of race and ethnicity in multivariate analyses), and the extent to which race and ethnicity were addressed in presentations of results, interpretation, and discussion. RESULTS: Relevant inclusion policies and regulations were identified from U.S. Agency for Healthcare Research and Quality (AHRQ), Agency for Toxic Substances and Disease Registry/ Centers for Disease Control (ATSDR/CDC), Food and Drug Administration (FDA), Health and Human Services (HHS), and National Institutes of Health (NIH). More than 80% of studies reviewed were supported at least in part by NIH or CDC, and race and ethnicity were mentioned in more than 90% of studies; however, fewer than half of these studies provided a rationale for including race and ethnicity data or the methods used for collecting such information. Three out of four studies included race and ethnicity in statistical analyses, yet the results of these analyses were discussed in only half of studies reviewed. CONCLUSION: Although reporting of race and ethnicity in the epidemiologic literature has increased in recent years, this review finds the treatment of race and ethnicity to be incomplete and inconsistent. The use of race and ethnicity in epidemiologic studies can be improved by implementing existing policy. doi:10.1016/j.annepidem.2004.07.047
P048S ASSOCIATION BETWEEN STATE MANDATE OF CONTRACEPTIVE INSURANCE COVERAGE AND INFANT MORTALITY IN THE UNITED STATES, 2001 EL Ding, TA LaVeist, Department of Health Policy and Management, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD PURPOSE: Previous research has demonstrated that more liberal abortion policies are associated with fewer unintended births and lower neonatal mortality rates. Studies also report a link between unintended pregnancies and adverse birth outcomes such as low birth weight, a strong predictor of infant mortality. This study seeks to examine the effect of state legislative policies mandating health insurance coverage for contraceptives and its association with infant mortality. METHODS: National Center for Health Statistics data of 2001 neonatal mortality rates (NMR), infant mortality rates (IMR), and percent births of low birth weight (LBW) were compiled for U.S. states. State policies mandating insurance coverage for contraceptives were ranked as ‘‘none, limited, or comprehensive’’ by the Kaiser Family Foundation. Analyses regressed ‘‘any-coverage-
609
mandate’’ with infant health indicators, adjusting for potential confounders. Stratification for race adjusted for race-respective covariates. RESULTS: In 2001, 26 states possessed a mandate for contraceptive coverage. In crude analyses, any-coverage-mandate was associated with overall lower NMR (b Z ÿ0.63, P Z 0.04) and lower IMR (b Z ÿ1.25, P Z 0.003). In multivariate analyses, though negative regression coefficients for LBW and NMR were not significant, marginally lower IMR (b Z ÿ0.73, P Z 0.06) was found among states with any-coverage-mandates. When stratified by race, no associations were found among whites between coverage-mandate and LBW, NMR, and IMR. Among blacks, however, coverage-mandate was significantly associated with lower LBW percentage (b Z ÿ0.78, P Z 0.04), lower NMR (b Z ÿ2.78, P Z 0.01), and lower IMR (b Z ÿ3.76, P Z 0.02). CONCLUSION: State policies mandating contraceptive insurance coverage appear correlated with fewer adverse birth outcomes, especially among blacks. This corroborates the hypothesis that state policy mandating insurance coverage for contraceptives may help prevent unintended births and lower subsequent infant mortality rates by facilitating greater access of contraceptives to women of lower SES. doi:10.1016/j.annepidem.2004.07.048
P049 ADDITION OF AN ACTIVE SURVEILLANCE COMPONENT TO INFLUENZA-LIKE ILLNESS SURVEILLANCE IN PENNSYLVANIA, 2003–2004 JS Samkoff, M Brookes, Division of Infectious Disease Epidemiology and Division of Immunization, Pennsylvania Department of Health PURPOSE: The Pennsylvania Department of Health (PADOH) has participated in the CDC Influenza Surveillance Provider Network (ISPN) since 1997. In 2003, we added an active surveillance component in an effort to improve the consistency with which providers submit weekly reports of influenza-like illness (ILI) seen in their practices. Because active surveillance entailed nontrivial costs for PADOH, some measurement of its benefits was necessary to facilitate policy decisions regarding use of active ILI surveillance in future influenza seasons. METHODS: Each Tuesday from November 3, 2003, through March 2, 2004 (CDC weeks 200345 through 200409), a list of ISPN providers who did not submit an ILI report for the previous week was generated from the CDC ISPN website. Immunization nurses in the local health districts phoned these providers and offered assistance with ILI data entry. Primary measure: the difference in the percentage of ISPN providers submitting reports in 50% or more of CDC-defined influenza surveillance weeks in the 2002–2003 influenza surveillance season (no active surveillance) and the 2003–2004 season (with active surveillance). Subsidiary measure: change in reporting consistency for individual providers who participated in ISPN during both seasons, as measured by the mean and median change in % of surveillance weeks for which providers submitted reports. RESULTS: Forty-two sentinel providers participated in the entire 2003–2004 surveillance season and 66 in 2002–2003. Active surveillance resulted in an absolute increase of 46% (from 23% to 69%) in proportion of ISPN providers reporting in 50% or more of
610
ABSTRACTS (ACE)
AEP Vol. 14, No. 8 September 2004 : 592–624
surveillance weeks. Among the 37 providers who participated during both seasons, reporting consistency improved a mean of 31% (95% CI Z 20%, 43%), median 30%. CONCLUSION: Active surveillance was a valuable addition to Pennsylvania’s ISPN. Consideration of the improvement achieved versus cost incurred will be incorporated into decision making regarding future use of this active surveillance technique by PADOH. doi:10.1016/j.annepidem.2004.07.049
P050S ESTABLISHING SURGICAL POLICIES IN EAST AFRICA: RESULTS OF THE EASI-DELPHI PROJECT Massey Beveridge, Kirsteen Burton, Office of International Surgery, University of Toronto, Ontario, Canada PURPOSE: In East Africa, demand for surgical services is high: 400 surgeons provide care to a population of more than 200 million people. Given the paucity of official surgical policies in the region and the need for systematic regional planning to improve the delivery of surgical services in East Africa, as a start, basic surgical priorities needed to be identified. METHODS: Thirty-one members of the Association of Surgeons of East Africa (ASEA) participated in a Delphi process. The first stage consisted of a survey of surgical issues in East Africa. The results of the survey were circulated to the participants who were then asked to generate statements in response to the question, ‘‘What actions will most reduce the burden of surgical disease in East Africa by 2010?’’ Seventy-nine statements were received; after combining similar statements, 60 were returned to the group, who then scored the desirability and feasibility of each statement. Lower-scoring items were discarded and the remaining 25 statements were returned to the participants for ranking. The 10 statements with the highest mean scores and least variance were identified. RESULTS: Five of the top 10 priorities identified were to (i) improve opportunities for continuing medical education for practicing surgeons, (ii) introduce more surgical skills workshops for medical students and clinical officers, (iii) involve COSECSA in surgical training as well as curriculum development and certification of surgeons, (iv) provide a feedback system by which medical students and surgical trainees may evaluate their teachers, and (v) recruit and train more nurses and anesthetists. CONCLUSION: If adopted and implemented, these priorities may help the ASEA, African Ministries of Health, surgical educators, hospital administrators and individual surgeons to reduce the burden of surgical disease in East Africa by the year 2010. doi:10.1016/j.annepidem.2004.07.050
HEALTH SERVICES P051S WOMEN AND WILLINGNESS TO PARTICIPATE IN CLINICAL TRIALS: RESULTS FROM A HYPOTHETICAL RANDOMIZED CONTROL TRIAL EL Ding, NR Powe, JB Braunstein, Johns Hopkins Medical Institutions, Baltimore, MD
PURPOSE: The U.S. Congress, the Institute of Medicine, and the National Institutes of Health have all expressed great concern about insufficient representation of women in medical research, particularly within large-scale cohort and clinical safety and efficacy trials. Despite federal mandates requiring adequate representation and enrollment of women in such healthcare research, gender-related participation disparities often persist. We asked whether underrepresentation of women in clinical trials could be due to a lesser willingness to participate (WTP) on the part of women relative to men. METHODS: We randomly approached patients from 13 Maryland outpatient clinics to self-complete a survey of their WTP in a hypothetical cardiovascular chemoprevention randomized control trial (RCT). After participants rated their WTP on a five-point Likert scale (CWTP Z very likely, likely), impact of gender on WTP was analyzed from logistic regressions, adjusted for potential sociodemographic and clinical confounders. Interactions between gender and covariates were also evaluated. RESULTS: Of 1132 eligible individuals, 70% responded. By gender, 31% of women were WTP in the hypothetical RCT, compared with 38% of men. In crude analysis, women were marginally less willing to participate in the trial (odds ratio OR Z 0.74, 95% confidence interval CI Z 0.55–1.00, P Z 0.049). In the adjusted model, women indicated a significantly lower WTP than men (OR Z 0.61, 95% CI Z 0.49–0.77, P ! 0.001). Covariates interacting with gender associated with increased female WTP included lower education (!12 years education; P Z 0.001), having a sick family member (P Z 0.03), being current smoker (P Z 0.08), Caucasian race (P Z 0.10), having greater income (O$30,000/yr; P Z 0.09), and rating religion as being less important in their lives (P Z 0.04). CONCLUSION: Women appear less willing to participate in randomized control trials than men. This gender disparity in trial participation may be an obstacle to federal mandates requiring adequate representation of women in cohorts and clinical trials. doi:10.1016/j.annepidem.2004.07.051
P052 DIRECT MEDICAL COSTS OF PRETERM BIRTH FROM BIRTH TO AGE 7 YEARS: A POPULATION-BASED STUDY IN MANITOBA, CANADA CV Newburn-Cook, M Heaman, D Schopflocher, E Forget, P Jacobs, O Casiro, J Blanchard, R Sauve, LL Roos, Universities of Alberta and Manitoba, Canada PURPOSE: (i) To determine direct medical costs by gestational age and survival status from birth to age 7 years in two cohorts (16,079 newborns in 1987 and 15,853 newborns in 1993) in Manitoba, Canada, and (ii) to determine the incremental costs associated with preterm birth. The 1993 cohort was chosen to examine the impact of surfactant on survival and costs. METHODS: A longitudinal population-based study using administrative data to calculate actual costs of hospital inpatient care (based on Refined Diagnostically Related Groups), physician and lab or diagnostic costs associated with hospital stays, and outpatient care, using 1995 constant dollars. RESULTS: Costs were inversely associated with gestational age. Substantial costs were associated with the initial hospital stay and first year of life; costs declined dramatically over the next 6 years.