Are Women’s Health Concerns Prioritized at the NIH and the FDA? Nicole C. Quon, Ph.D. Assistant Professor Indiana University
Scientific Agencies Scientific agencies use scientists and scientific evidence to make science policy Scientific agencies use scientists and scientific evidence to make science policy Likely to seek bureaucratic autonomy Likely to seek bureaucratic autonomy May respond to external pressure under certain conditions May respond to external pressure under certain conditions
Women’s Health Movement Relied on frames of gender inequity Relied on frames of gender inequity Concerns about medical research Concerns about medical research –Increasing attention to women’s health –Participation of women in clinical trials –Research funding for women’s health issues, especially breast cancer
Mobilizing Resources MeanRange General women’s health groups – 14 Disease-specific women’s health groups – 9 Other disease-specific groups – 56 National Women’s Health Network budget ($1000) – 743 Scientific agencies may consider the demands of resource-rich groups Scientific agencies may consider the demands of resource-rich groups
Raising Awareness MeanRange Congressional oversight index (alpha=0.62) – 42 Media coverage index (alpha=0.91) Scientific journal articles – 3467 Scientific agencies may respond to signals of issue importance Scientific agencies may respond to signals of issue importance Political/social influence vs. scientific influence Political/social influence vs. scientific influence
Reducing Monitoring Costs MeanRange Women in senior NIH positions – 9 Women in NIH study sections (% change) – 5.2 Administrative procedures Administrative procedures –Introduce decision makers who share the same values
Gender Politics MeanRange Women in Congressional committees (%) – 16.7 Positive gender gap in Presidential election (%) – 11 Negative gender gap in Presidential election (%) – 0 Partisanship in Congress – 1.4 Issues related to gender may become more salient under certain conditions Issues related to gender may become more salient under certain conditions
Disease Burden MeanRange Hospital admission rate for women – 1960 Overall death rate for men – 109 Prevalence rate for women – 7660 Disease-specific death rate for men – 42 Agency missions reflect public health goals Agency missions reflect public health goals Rate for women or men could influence priority setting Rate for women or men could influence priority setting
NIH Dependent Measures Related to decisions in the NIH grants program Related to decisions in the NIH grants program –Grants for “women or female” studies Extramural program (n=556) Intramural program (n=418) –Grants for studies on 23 diseases on the women’s health agenda Extramural program (n=749) Intramural program (n=660) Collected from the CRISP database of funded grants from 1972 to 2004 Collected from the CRISP database of funded grants from 1972 to 2004 Keyword searches of grant titles and abstracts Keyword searches of grant titles and abstracts
NIH Independent Variables Mobilization of resources Mobilization of resources Raising awareness Raising awareness Reducing monitoring costs Reducing monitoring costs Gender politics Gender politics Disease burden Disease burden Other variables Other variables –Female medical school faculty, year trend, presidential dummies
NIH Model Specification Count data Count data –Data was overdispersed –Data was a panel design 32 years 23 institutes or 23 diseases Random effects negative binomial models Random effects negative binomial models Offset to account for varying institute sizes Offset to account for varying institute sizes Lagged independent variables Lagged independent variables
NIH Extramural Priorities Models Studies on Women or Females Studies on 23 Disease Priorities Coeff.S.E.Coeff.S.E. Mobilizing Resources General women’s health groups * Disease-specific women’s health groups *** Other disease-specific groups *** National Women’s Health Network budget Raising Awareness Congressional oversight On women’s health On women’s health *** On specific disease On specific disease0.0132** Media coverage *** Scientific journal articles *** * p<0.10, ** p<0.05, *** p<0.01
NIH Extramural Priorities Models Studies on Women or Females Studies on 23 Disease Priorities Coeff.S.E.Coeff.S.E. Reducing Monitoring Costs Women in senior NIH positions * Women in NIH study sections Political Salience Women in Congressional committees ** Positive gender gap in Presidential election * *** Negative gender gap in Presidential election * Partisanship in Congress * p<0.10, ** p<0.05, *** p<0.01
NIH Extramural Priorities Models Studies on Women or Females Studies on 23 Disease Priorities Coeff.S.E.Coeff.S.E. Disease Burden Hospital admission rate for women Hospital admission rate for men Overall death rate for women Overall death rate for men Prevalence rate for women *** Prevalence rate for men Disease-specific death rate for women *** Disease-specific death rate for men *** * p<0.10, ** p<0.05, *** p<0.01
NIH Intramural Priorities Models Fewer influences seem to matter compared to extramural program decisions Fewer influences seem to matter compared to extramural program decisions Studies on women or females Studies on women or females –Gender politics: negative gender gap Studies on 23 disease priorities Studies on 23 disease priorities –Mobilizing resources: other disease-specific groups –Raising awareness: congressional oversight on specific diseases –Disease burden: death rate for men
FDA Dependent Measures Related to decisions for new drug approval Related to decisions for new drug approval –Assignment of “priority” review –Speed of new drug review in months Approval dates from 1970 to 2004 Approval dates from 1970 to 2004 Focused on drugs approved for diseases on the women’s health agenda (n=131) Focused on drugs approved for diseases on the women’s health agenda (n=131)
FDA Independent Measures Mobilizing of resources Mobilizing of resources –Interest groups Raising awareness Raising awareness –Congressional oversight, media coverage, scientific articles Disease burden Disease burden Other variables Other variables –FDA workload, previous firm success, PDUFA
FDA Model Specification Logistic regression to examine assignment of priority review Logistic regression to examine assignment of priority review Proportional hazards regression to examine the speed of drug review Proportional hazards regression to examine the speed of drug review
FDA Priorities Models Priority Review Drug Review Times Coeff.S.E.Coeff.S.E. Mobilizing Resources Disease-specific women’s health groups Other disease-specific groups Raising Awareness Congressional oversight on specific diseases Media coverage Scientific journal articles *** * p<0.10, ** p<0.05, *** p<0.01
FDA Priorities Models Priority Review Drug Review Times Coeff.S.E.Coeff.S.E. Disease Burden Prevalence rate for women Prevalence rate for men Disease-specific death rate for women Disease-specific death rate for men * Priority rating *** Other Variables FDA workload Previous firm success PDUFA trend * p<0.10, ** p<0.05, *** p<0.01
Summary of Main Results The FDA was responsive to the women’s health movement The FDA was responsive to the women’s health movement But not in priority setting for new drug approval But not in priority setting for new drug approval Female leadership (scientific and political) are associated with increased priority setting at the NIH Female leadership (scientific and political) are associated with increased priority setting at the NIH Congressional oversight and some signals from health advocates are also important Congressional oversight and some signals from health advocates are also important
Study Limitations NIH dependent measures collected using keywords NIH dependent measures collected using keywords Data on grant applications unavailable Data on grant applications unavailable Women’s health advocacy measure is crude Women’s health advocacy measure is crude Few drugs for diseases on the women’s health agenda Few drugs for diseases on the women’s health agenda
Policy Implications Scientific agencies are not insulated from gender politics Scientific agencies are not insulated from gender politics Influence depends on the type of decision and agency culture Influence depends on the type of decision and agency culture Some pathways of influence seem more effective Some pathways of influence seem more effective –Collaborations between interest groups and Congress –Increasing the role of women leaders
Pathways of External Influence “External signals” theory “External signals” theory –Josckow, Olson –Mobilizing resources –Raising awareness “Political control” theory “Political control” theory –Weingast and Moran, McNollGast, McCubbins and Schwartz –Reducing monitoring costs Political salience Political salience
Agency Mandates and Culture Research scientific agencies Research scientific agencies –NIH intramural grants program Distributive scientific agencies Distributive scientific agencies –NIH extramural grants program Regulatory scientific agencies Regulatory scientific agencies –FDA Center for Drug Evaluation and Research
NIH Independent Variable Lags Agencies respond to most recently available information Agencies respond to most recently available information 1 year lag: Congressional oversight, media, and scientific journal coverage 1 year lag: Congressional oversight, media, and scientific journal coverage 2 year lag: interest groups 2 year lag: interest groups 3 year lag: disease burden 3 year lag: disease burden
Grants for Women's Health Agenda Diseases Percent of Total NIH Grants Year
Studies on Women or Females (%) % 1-5 % 5-10 % >10%
Studies on Women or Females (%) % 1-5 % 5-10 % >10%
NIH Results Summary Priority setting in the NIH extramural and intramural programs for women’s health is not insulated from politics Priority setting in the NIH extramural and intramural programs for women’s health is not insulated from politics All four pathways of external influence seem to matter All four pathways of external influence seem to matter Extramural decisions are associated with more external influences Extramural decisions are associated with more external influences
Priority Review of New Drugs
Mean Drug Review Times (in months)
FDA Independent Measures I MeanRange Mobilizing Resources Disease-specific women’s health groups – 10 Other disease-specific groups – 46 Raising Awareness Congressional oversight index – 16 Media coverage index – 923 Scientific journal articles – 12097
FDA Independent Measures II MeanRange Disease burden per 10,000 population Prevalence rate for women – 4498 Prevalence rate for men – 5590 Death rate for women – 30 Death rate for men – 40 Other Variables FDA workload % of firms with previous success 73%
Directions for Future Research Examine impact of women’s and women’s health movement on other scientific agencies Examine impact of women’s and women’s health movement on other scientific agencies Study whether other disease groups that do not have historical gender inequities have influenced scientific agencies decisions Study whether other disease groups that do not have historical gender inequities have influenced scientific agencies decisions