Variations in breast and cervical cancer screening prevalence among women veterans

Goldzweig C, Caffrey C, Rubenstein LV, Young GL, Yano EM. Variations in breast and cervical cancer screening prevalence among women veterans. Paper presented at: VA HSR&D National Meeting; 2001 Feb 15; Washington, DC.

Objectives: Increased emphasis on prevention is central to primary care for all veteran users, however, women veterans are a minority in the veteran population and little is known about the organizational characteristics that promote gender-appropriate screenings for women veterans. The purpose of this work is to examine the features of VA medical centers that are associated with guideline-adherent breast and cervical cancer screening. Methods: We linked site-specific results from the VHA Primary Care Delivery Models Survey (1996) administered as a mailout survey to primary care directors at geographically distinct VA facilities (100% response rate) to site-specific results from the VA External Peer Review Program (EPRP) chart reviews on indicated prevention activities (i.e., Pap smears, mammography) based on recommendations adapted from the National Preventive Services Task Force (n = 145). We integrated site-specific organizational culture and quality improvement orientation scores from the Management Decision and Research Center. We conducted bivariate and multivariate analyses of prevention performance (proportion guideline-adherent) by key environmental (e.g., region), organizational (e.g., complexity) and primary care (e.g., firms) features to describe variations in practice. Results: We found no differences in performance of mammography or Paps by region, location, facility complexity, or academic affiliation in contrast to generic prevention measures (e.g., flu shots), where complexity and affiliation are significant independent predictors of poorer performance. The prevalence of Pap smears was higher in VA's with more general internal medicine doctors (p < .05), and where primary care providers were promptly notified of subspecialty consultation results (p < .05). However, performance did not vary by self-reported level of primary care implementation, level of organized primary care (e.g., firms) or by other key primary care features. Cervical cancer screening rates were higher in VA's with higher quality improvement orientation (p < .05), with higher leadership ratings (p < .05) and higher reward and recognition scores (p < .005), with a trend toward having more primary care process action teams (p = .08). Mammography use did not significantly vary by any of these local characteristics, with the exception of higher rates in VA's with higher reward and recognition scores (p < .05). Conclusions: Preventive care for women veterans does not vary by the same types of care features as found in other structure-outcome models that typically reflect male veterans because of their preponderance among sampled users. Women veterans either do not benefit from the influence of organized primary care in the same manner as male veterans or do not use primary care in the same way (i.e., use women's clinics or divide care between VA and non-VA sites more than male veterans). Impact: The proportion of women veterans across VA facilities varies and may pose logistical barriers to assuring access to preventive care, especially when resources are constrained. The types of delivery models or care features to which women veterans are exposed is as yet unclear, as is the effectiveness of alternate models (e.g., integrated primary care vs. separate women's health clinic). Further research is needed to ascertain the structure of care that women veterans experience and prefer, as well as the effectiveness of alternate structures on outcomes.