SOLICITATION NOTICE
R -- BIOSTATISCAL SERVICES
- Notice Date
- 8/22/2007
- Notice Type
- Solicitation Notice
- NAICS
- 541690
— Other Scientific and Technical Consulting Services
- Contracting Office
- Department of Health and Human Services, National Institutes of Health, Nat'l Institute of Diabetes, Digestive, & Kidney Diseases, 2 Democracy Plaza, Suite 700W 6707 Democracy Blvd., MSC 5455, Bethesda, MD, 20892-5455, UNITED STATES
- ZIP Code
- 00000
- Solicitation Number
- NIH-NICHD-07-370
- Response Due
- 9/6/2007
- Archive Date
- 9/21/2007
- Description
- The National Institutes of Health (NIH), National Institute of Child Health and Human Development (NICHD), Perinatology Research Branch (PRB) intends to procure the services of David Todem to support the research effort of the PRB of the NICHD through design and analysis of longitudinal studies using comprehensive, state of the art statistical methods. The contract shall provide study investigators with expert statistical advice to be used in all phases of conducting individual studies. This is a continuation of statistical and methodological methods developed by David Todem in Perinatal Health for PRB. Specific tasks include the following: The Contractor shall utilize the existing databases, within the Branch, that consist of previously collected obstetric data, to facilitate investigations of preeclampsia, HELLP syndrome and small neonates for gestational age. In particular, the Contractor shall consider both parametric, semiparametric and nonparametric statistical models needed to understand the relationship between pregnancy outcomes (Preeclampsia, Small neonates for gestational age, and uncomplicated pregnancies) and prenatal biologic analytes (VEGFR-1, PIGF, and s-eng Endoglin) measured over time. Contractor shall use models that appropriately account for association in the data. Parametric models will include both linear and nonlinear models. Specifically, the Contractor shall use parametric growth curve models (mixed-effects models, nonlinear mixed effects models and population-averaged models) to prenatal biologic analytes viewed as responses. To increase efficiency, the Contractor shall consider model-based standard errors after a careful modeling of the association structure in the data. Other parametric model shall include prediction models based on polytomous logistic regression for identifying patients with the clinical disease, using either each analyte, and as well as a combination of analytes viewed as covariates. Receiver operating characteristic curves (ROC) shall also be constructed and classification matrices generated. The contractor shall use semiparametric models that use robust ?Sandwich? estimators to estimate the standard errors of the parameter estimates when the association structure in the data is misspecified. Nonparametric models will include descriptive tools such as spline models and explanatory tools such as regression trees and classification trees for longitudinal data. It is common that in medical research, with the current technology, a level of some analytes may be detectable only above a threshold level. A good example in the database available at the Branch is the PIGF analyte, which is only measured above some threshold value. The Contractor shall then consider a piece-wise growth curve model or any other novel tool to adequately capture this feature of the data. The Contractor shall check all model fits for individual and systematic departures of the observed and fitted values using informal (e.g. inspection of residuals) and formal methods (e.g. based on Wald, Likelihood ratio and score tests for extra parameters). If individual outliers are detected, their influence will be evaluated by comparing models with and without outlying values. When we are not successful in fitting the parametric model, then non-parametric methods such as generalized additive models shall be utilized. To minimize bias in the prediction model, the Contractor shall consider a cross-validation method by splitting the original data into a training sample (about ? of data) and a testing set (about ?). Contractor will provide the background and reasoning for using a particular method of analyzing the data. Contractor will provide the PRB with references for the method(s) chosen. Contractor shall utilize updated versions of software packages that he/she considers most appropriate for the type of data handled. In addition the Contractor shall consider programming new methodologies and models using matrix oriented-software packages as these new models may not be fitted using commercial software packages. Contractor will perform intensive computation, including bootstrapping and permutation, for statistical hypothesis testing. The Contractor shall provide administrative support for the research projects such as typing of statistical reports and submissions for publications; literature searches and collection of scientific papers; copying of scientific papers; maintaining and updating a computerized reference library (e.g. EndNote). Place of Performance and additional tasks Work is to be performed by Contractor both at the PRB site at Wayne State University located in Detroit, Michigan and at Contractor?s own location. To function effectively, the PRB will hold frequent meetings input from, and collaboration with, its Cooperative Group partners and other stakeholders. The Contractor shall attend in-person onsite month meetings organized by the head of PRB or his designee at the PRB premises, for an overview of the conduct of the project. In addition, the Contractor shall organize and coordinate in-person weekly meetings (preferably on Fridays) on methodological and statistical issues of the project in collaboration with the PRB project director or designee. The Contractor should be a member of a task specific panel made of all consultants that offer statistical advice and reviews about all aspects of PRB statistical operations. Areas in which this task-specific panel would be of value include, study design, data management and analysis, and interpretation of results. This panel will primarily meet every week in person, although the PRB may elect, in addition, to host conference call meetings to minimize costs for in-person meetings. Deliverables Contractor will deliver the information in an understandable format including tabular and graphic presentations of the data. Contractor will make formal presentations of the data analyses and results. The Contractor shall also co-author, in collaboration with the PO, other staff members of PRB as appropriate, the results of these investigations at appropriate scientific meetings and in peer-reviewed scientific journals. The Contractor will keep the PO appraised of research and anticipated publications and presentations on an ongoing basis. All anticipated papers, abstracts, and other publications will go through the usual review procedures of PRB prior to submission. Because of the nature of the work done in the contract, it is expected that expect under unusual circumstances, statistical and methodological publications will include one or more PRB staff as co-authors. By Authority: 41 U.S.C. 253 (c) (1), as set forth in FAR 6.302-1, only one responsible source and no other supplies or service will satisfy agency requirements. Interested vendors may identify in writing their interest with a capability statement to the Contracting Office prior to the closing date of this notice. This notice of intent is not a request for competitive proposals, however, all responses received, before the closing date, of this announcement will be considered by the Government. A determination by the Government not to compete this proposed contract is based upon responses to this notice and is solely within the discretion of the Government. Information received will be considered solely for the purpose of determining whether to conduct a competitive procurement. Inquiries may be directed to Tina Robinson via email: robinsti@mail.nih.gov
- Record
- SN01380828-W 20070824/070822220616 (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 |