Developing recommendations for evidence-based clinical preventive services for diverse populations: Methods of the u.s. Preventive services task force

Annals of Internal Medicine R ESEARCH AND R EPORTING M ETHODS Developing Recommendations for Evidence-Based Clinical Preventive Services for Diverse Populations: Methods of the U.S. Preventive Services Task Force Kirsten Bibbins-Domingo, PhD, MD, MAS; Evelyn Whitlock, MD, MPH; Tracy Wolff, MD, MPH; Quyen Ngo-Metzger, MD, MPH; William R. Phillips, MD, MPH; Karina W. Davidson, PhD, MASc; Alex H. Krist, MD, MPH; Jennifer S. Lin, MD, MCR; Carol M. Mangione, MD, MSPH; Ann E. Kurth, PhD, CNM, MSN, MPH; Francisco A.R. Garci´a, MD, MPH; Susan J. Curry, PhD; David C. Grossman, MD, MPH; C. Seth Landefeld, MD; John W. Epling Jr., MD, MSEd; and Albert L. Siu, MD, MSPH The U.S. Preventive Services Task Force (USPSTF) summarizes the principles and considerations that guide development of its recommendations for diverse U.S. populations. It uses these principles through each step in the evidence-based guideline process: developing the research plan, conducting the evidence review, developing the recommendation, and communicating to guideline users. Three recent recommendations provide exam- ples of how the USPSTF has used these principles: the 2015 recommendation on screening for abnormal blood glucose and type 2 diabetes; the 2016 recommendation on screening for breast cancer; and the recommendation on screening for pros- tate cancer, which is currently in progress. A more comprehen- sive list of recommendations that includes considerations for specific populations is also provided. T proportionately affected by a condition or susceptible to variation in the effectiveness of the preventive ser- vice. In consultation with the USPSTF, the EPC investi- gators have developed a comprehensive approach to incorporating the evidence for diverse populations throughout all phases of the systematic review process, including determination of the scope of the topic, data abstraction and critical appraisal, data analysis and syn- thesis, and reporting and interpretation of the evidence (13). Additional input about subpopulations comes from outside review and public comment on the draft research plan to further refine our process. For exam- ple, the draft and final research plans on screening for prostate cancer (Supplement 1, available at Annals.org) highlight additional clarifications from the USPSTF on consideration of African American men and men with family history of prostate cancer, which were added in response to public comments. The research plan guides the systematic evidence review conducted by the EPC team, and the resulting evidence report routinely includes information on the epidemiology across all relevant populations (for exam- ple, incidence, prevalence, and mortality). Supplement 2 (available at Annals.org) shows the draft and final research plans for the 2015 recommen- dation on screening for abnormal blood glucose and type 2 diabetes. The analytic framework depicts the intent to review literature and consider variation in benefits and harms by risk status (high vs. average) and to examine additional variability by age, sex, and race/ ethnicity at each stage in the framework. The subse- he U.S. Preventive Services Task Force (USPSTF) works to improve the health of all Americans by making evidence-based recommendations about clini- cal preventive services, such as screening, counseling, and use of preventive medications (1). To achieve this goal, we pay particular attention to how our recom- mendations can be most effectively applied to specific segments of the U.S. population with patterns of disease or the effectiveness of a preventive service that may differ from the general population. These specific populations may be identified by demographic charac- teristics (for example, age, race/ethnicity, or sex) or other factors (for example, biology, behavior, or heredity). In this article, we outline the principles and consid- erations that guide the development of our recommen- dations for specific U.S. populations. We discuss 3 recent recommendations as examples: the 2015 rec- ommendation on screening for abnormal blood glu- cose and type 2 diabetes (2); the 2016 recommenda- tion on screening for breast cancer (3); and the recommendation on screening for prostate cancer, which is currently in progress (4). A more comprehen- sive list of recommendations that includes consider- ations for specific populations is provided in the Table. D EVELOPING R ESEARCH P LANS AND C ONDUCTING THE E VIDENCE R EVIEW TO I NCLUDE D IVERSE P OPULATIONS We consider diverse populations starting with the first step in our process— development of the research plan. This plan defines the types of evidence that will be gathered and reviewed by the Evidence-based Prac- tice Center (EPC) team and used by the USPSTF to de- velop the recommendation. The research plan rou- tinely includes a means to identify evidence on whether specific segments of the U.S. population may be dis- Ann Intern Med. 2017;166:565-571. doi:10.7326/M16-2656 For author affiliations, see end of text. This article was published at Annals.org on 7 March 2017. Annals.org See also: Web-Only Supplement Annals.org Downloaded From: http://annals.org/ by a UCLA Digital Collection Svs User on 09/07/2017 Annals of Internal Medicine • Vol. 166 No. 8 • 18 April 2017 565