SOURCES SOUGHT
B -- Average Treatment Effect Estimation Accounting Covariate Measurement Error New Causal Model Technique with a Medical Device Application: The Impact of Breast Pump Use on Mother's Breast-Feeding Practice and Infant Health
- Notice Date
- 6/11/2010
- Notice Type
- Sources Sought
- NAICS
- 541690
— Other Scientific and Technical Consulting Services
- Contracting Office
- Department of Health and Human Services, Food and Drug Administration, Office of Acquisitions and Grants Services, 5630 Fishers Lane, Room 2129, Rockville, Maryland, 20857-0001
- ZIP Code
- 20857-0001
- Solicitation Number
- REQ1073527
- Archive Date
- 7/2/2010
- Point of Contact
- Mary Rose A. Nicol, Phone: 3018277183
- E-Mail Address
-
MaryRose.Nicol@fda.hhs.gov
(MaryRose.Nicol@fda.hhs.gov)
- Small Business Set-Aside
- Total Small Business
- Description
- THIS IS A SOURCES SOUGHT NOTICE to determine the availability and capability of Small Business Firms. This notice is for planning purposes only, and does not constitute an Invitation for Bids, a Request for Proposals, Solicitation, Request for Quotes, or an indication the Government will contract for the items contained herein. This notice is not to be construed as a commitment on the part of the Government to award a contract, nor does the Government intend to pay for any information submitted as a result of this notice. The Government does not reimburse respondents for any cost associated with submission of the information being requested or reimburse expenses incurred to interested parties for responses to this sources sought. Any responses received will not be used as a proposal Title: Average Treatment Effect Estimation Accounting for Covariate Measurement Error - New causal model technique with a medical device application: the Impact of Breast Pump Use on Mother's Breast-Feeding Practice and Infant's Health Background From a methodological point of view, ignoring measurement error in covariates (a fairly common issue) may lead to misleading inferences on average treatment or exposure effects evaluation in non-randomized studies. Although an analysis based on a carefully conducted, randomized and controlled clinical trial is still the gold standard in obtaining valid causal effects of medical products, such designs can be either impractical or too burdensome to conduct in pre-market and post-market studies. For example, for medical devices, it is not uncommon to have a non-randomized design for the Investigational Device Exemption (IDE) study in the premarket phase; and very frequently a prospective, controlled cohort design is used for the Post-approval Study (PAS) of medical devices at the post-market phase. Average causal effect (ACE) estimation methods for these non-randomized studies have typically relied on standard propensity scoring techniques under the traditional causal framework, which assumes all covariates are measured accurately and no unobserved factors influence the treatment and outcome (Rosenbaum and Rubin, 1983,1984; Dehejia and Wahba, 1999). However, covariates are often measured with unobservable error, which violates the critical assumption underlying the current causal framework. Since ignoring the error often leads to significant biased results, extending the standard causal inference framework allowing covariate measurement error and developing appropriate estimation methods are of critical importance for a valid assessment of the safety and efficacy of medical products (including devices, drugs, and biologics) in both the pre-market and post-market phases. From an application point of view, the breast pump is a medical device regulated by CDRH and broadly used in US. A study has shown that 85% of mothers who breastfed infants that were 1.5-4.5 months old used a breast pump. However, little is known about how mothers' experiences with breast pumps, including both benefits and adverse events, impact their breast feeding practice. The American Academy of Pediatrics recommends that infants receive breast milk throughout the first year of their life. Therefore, it is important to know how the mother's breast feeding practice associated with the use of a breast pump and how the breast pump usage influences the infant's health. However, some important confounders, e.g. household social economic status and mother's post-natal health status, are either unobservable or liable to measurement error. The presence of measurement error in covariates violates the critical assumption underlying the current causal framework, which need new causal methods to address these two important questions. In this project, we plan to use IFPS II data (Infant Feeding Practices Study II) (Fein, 2008), a major longitudinal study conducted by FDA in collaboration with CDC, NIH, DHHS/OWH, and the Maternal and Child Health Bureau. The data have been analyzed for presentation at the annual meetings of the American Public Health Association in November 2007 and were published in a journal supplement in 2008. Purpose The purpose of this requisition is to procure appropriate source to: (1) develop a causal inference framework to allow covariate measurement error; (2) develop a new causal inference model for average treatment effect estimation accounting for the measurement error structure under non-randomized clinical trials and observational studies; and (3) apply this new causal model on the evaluation of the health impact of breast pump usage using IFPS II data. Scope of Work The contractor shall: • develop a new statistical methodology to: o create a causal inference framework to allow for covariate measurement error. o develop an average treatment effect (ATE) estimation model under the new causal framework that accounts for covariate measurement error, based on existing propensity score methodology. o conduct simulation studies to demonstrate the advantages of the new method over the currently used propensity score-based causal inference method. • participate in developing and writing manuscripts describing the extended causal framework and the proposed new estimation methods. • conduct the data analysis, using IFPS II data, to evaluate the health impact of breast pump usage on mother's breast feeding behavior and infants' health status using the newly developed causal inference model and conventional propensity score-based causal inference method. • participate in writing manuscripts describing the impact of breast pump use on Mother and infant health. • submit the final report/documentation of the developed new Causal Inference Methodology. Timeline with Major Milestones and Deliverables The contractor shall submit the following: (all submissions shall be electronic) 1) Statistical plan of developing a new causal inference methodology to allow covariate measurement error, and the documentation for the development of estimation methods. 2) Manuscript for the development of average treatment effect (ATE) model under the proposed causal framework accounting for covariate measurement error. 3) Data analysis results for the evaluation of the impact of breast pump usage on mother and infant health using IFPS II data. 4) The SAS or statistical programs used to analyze the data, and an interpretation of the results. 5) Final report and documentation of new Causal Inference Methodology and its application on the study of Breast Pump Use. Schedule of Deliverables Scheduled items Due Dates: 1) Statistical plan of developing a new causal inference methodology to allow covariate measurement error, and the documentation for the development of estimation methods 10-30-2010 2) Manuscript for ATE estimation - 12-31-2010 3) Data analysis results for the evaluation of breast pump usage - 03-30-2011 4) Interpretation of results for the drafting of manuscript - 05-30-2011 5) Final report and documentation of new Causal Inference Methodology and its application on the evaluation of Breast pump - Use 08-30-2011 Period of Performance May 1, 2010 to August 30, 2011. Place of Performance The work shall be performed at the contractor's site. Interested Small Business potential offerors are encouraged to respond to this notice. However, be advised that generic capability statements are not sufficient for effective evaluation of respondents' capacity and capability to perform the specific work as required. Responses must directly demonstrate the company's capability, experience, and/or ability to marshal resources to effectively and efficiently perform the task described above at a sufficient level of detail to allow definitive numerical evaluation; and evidence that the contractor can satisfy the minimum requirements listed above while in compliance with FAR 52.219-14 ("Limitations on Subcontracting"). Failure to definitively address each of these factors will result in a finding that respondent lacks capability to perform the work. Responses to this notice shall be limited to 10 pages, and must include: 1) Personnel: The offeror shall name the participating personnel and identify their qualifications and experience. The offeror shall provide personnel who have education in statistics, biostatistics, or a field that includes statistical training such as epidemiology. The personnel shall have formal training in longitudinal data analysis and causal inference methodology. 2) Relevant Experience: The offeror shall provide personnel who are familiar with the IFPS II study and dataset, have demonstrated professional knowledge of developing causal inference methodology modeling and longitudinal data analysis. 3. Name, title, telephone number, and e-mail addresses of individuals who can verify the demonstrated capabilities identified in the responses. 4) Project Management: The offeror shall provide information on the administration of the project. This should include management plans, methods for achieving data analysis and manuscript writing, and quality control procedures. 5) DUNS number, CAGE Code, Tax Identification Number (TIN), and company structure (Corporation, LLC, partnership, joint venture, etc). Companies also must be registered in the Central Contractor Registration (CCR) at www.ccr.gov to be considered as potential sources. 6) Identification of any GSA Schedule contract(s) by Schedule number and contract number and SINs that are applicable to this potential requirement are also requested. If the company has a Government approved accounting system, please identify the agency that approved the system. Please submit copies of any documentation, such as letters or certificates to indicate the firm's status (see item #3 above). To the maximum extent possible, please submit non-proprietary information. Any proprietary information submitted should be identified as such and will be properly protected from disclosure. Interested offerors should submit their capability statement not exceeding five (5) pages in length. Phone calls will not be accepted or returned. Interested firms or individuals may submit the requested information to: maryrose.nicol@fda.hhs.gov or US Food and Drug Administration Mary Rose A. Nicol 5630 Fishers Lane / HFA-500 OAGS/DAO Rm 2089 Rockville, MD 20857
- Web Link
-
FBO.gov Permalink
(https://www.fbo.gov/spg/HHS/FDA/DCASC/REQ1073527/listing.html)
- Record
- SN02175737-W 20100613/100611235124-8a29c2602fca419308b318488b0f4cd0 (fbodaily.com)
- Source
-
FedBizOpps Link to This Notice
(may not be valid after Archive Date)
| FSG Index | This Issue's Index | Today's FBO Daily Index Page |