Cytel Software Suite Renewal
Added: Aug 12, 2016 11:19 am Notice Type: Combine Synopsis/SolicitationOriginal Posted Date August 12, 2016Response Date: 8/22/2016Original Set Aside: Total Small businessClassification CodeNAICS 511210Synopsis: This is a combined synopsis/ solicitation for renewal of software in accordance with FAR Regulation (FAR) 12.102(g) and in the format of Subpart 12.6 and 13.5 as supplemented with additional information included in this notice.This announcement constitutes the only solicitation. Quotations are being requested and a written solicitation will not be issued. The solicitation number is FDA- SOL -16-1163455 and is issued as a Request for Proposals (RFP). This solicitation document incorporated all mandatory commercial item provision and clauses that are in effect tough the Federal Acquisition Circular 2005- 89 dated 7/14/2016.The North American Industrial Classification System is 511210 and the small size standard is 35M 500L This is a firm fixed price contract and is being issued as a "Brand Name or Equal to". If equal items are proposed, the quoter shall provide product specifications for those items proposed. All products proposed (whether brand name listed or equal to items proposed) must satisfy the functional requirements/salient characteristics set forth in this document. Title: Cytel Software Suite RenewalPeriod of performance: is from effective date of award tough 12 months plus one option year.A. BACKGROUNDStatisticians and other FDA personnel have a need for specialized software to evaluate, simulate, and analyze clinical trial designs, perform a wide variety of exact statistical analyses and calculations in support of regulatory science for both pre and post market reviews, perform power and sample size calculations for clinical trials, registries, and studies, as well as reliably evaluating designs for the emerging field of adaptive design trials. These specific and specialized capabilities cannot be found in general purpose software products such as SAS, R, Splus or SPSS. Cytel software has been used at the agency for the past 25 years, and is a defacto industry standard for exact statistical analyses, and clinical trial design and analyses, particularly in the emerging field of adaptive designs and group sequential designs.Failure to provide requested software or equivalent may result in the loss of ability to reproduce, verify, and validate sponsor results as well as conducting independent agency evaluations. This loss of capability may result in a failure of reviewers to meet agency deadlines for product reviews.B. OBJECTIVE OF THE PURCHASEThe objective of this acquisition is to obtain continuing and expanded access to specialized statistical analysis software for clinical trial design and evaluation, exact statistical analyses in both stand alone and SAS programming environments, power and sample size determination, and adaptive designs for phase I/II clinical trials and adaptive designs for phase II/III clinical trials. Additionally, this acquisition will provide for software maintenance, including upgrades and software patches, as well as technical support for the purchased software product.C. REQUIREMENTS/DELIVERABLESThe following Cytel products, or equivalent, are required:1. EAST 6.4, along EAST Adapt, EAST Survival, EAST Xact and SURVADAPT modules.2. EAST Endpoints, EAST Predict and EAST Escalate modules.3. StatXact PROCs4. StatXact 115. LogXact 116. Compass 2.0D.ResponcibilitiesThe vendor shall also provide licenses, support and software maintenance, including updates, for the software product. The vendor shall provide maintenance and support existing FDA perpetual licenses of StatXact and LogXact. The software must be capable of running on 64-bit Windows 7 systems.1. To be considered equivalent to the supplied software product must perform the following:Fixed Sample Size and Group Sequential Designs• Normal Designs o Superiority (one and two sample designs)o Non-inferiorityo Equivalence• Binomial Designs (including exact binomial tests)o Superiority (One, two and many samples)o Non-inferiority (two sample)o Equivalence (two sample)• Survivalo Superiority o Non-inferiority • Information Based Designs• Adaptive Designs and Simulationso Cui, Hung and Wang methods (Hung and Wang are in the Office of Biostatistics)o Mueller and Schafer binomial methodso Proshan and Hunsberger multi-stage designso Reanalysis with modified error spending function• Enhanced Survival Designs and Simulationso Superiority o Non-inferiority o allow for random drop outs and accrual rate specification functionso Proportional and non-proportional hazard alternativeso Non-uniform accrualo Study duration estimationo Accommodate differential drop out and varying hazard functionso Allow sample size re-estimation based on unblended interm analysis• Graphics and Chartso Boundaries o Error Spending functionso Power versus sample size• Sample Size Re-estimationo Normal o Binomialo Survival• Interim Monitoringo Nominal Critical Point o Error Spending o Conditional Powero Post-Hoc Power• Traditional rule-based algorithms (3+3)• Continuous Reassessment Methods (mCRM; Goodman et al. 1995)• Modified Toxicity Probability Model (mTPI; Ji et al. 2010)• Bayesian Logistic Regression Model (BLRM; Neuenschwander et al. 2008)• Predicted Interval Plots• Flexible Interim Computation: Accommodate unexpected changes in number of looks and their timing• Simulations of trial endpoints for feasibility studies during design stage• Accurate calculations of power and Type 1 error for ongoing trials• Updating methods for information about trial deadlines and milestones• Predictions of enrollment milestones• Predictions of patient recruitment and country performance• Multiple primary endpoints • Multiple primary and secondary endpoints • Serial gatekeeping• Parallel gatekeeping2. Power and Sample Size. A Power and sample size calculation software package/component which includes the following analyses and capabilities:Continuous:• Single Meano Student to Normal• Difference of Paired Meanso Superiority: Student to Superiority: Normalo NonInferiority: Student to NonInferiority: Normalo Equivalency: Bivariate Student t• Ratio of Paired Meanso Superiority: Student to Superiority: Normalo NonInferiority: Student to NonInferiority: Normalo Equivalency: Bivariate Student t• Wilcoxon Signed Rank TestTwo Means:• Difference of Independent Meanso Superiority: Student to Superiority: Satterthwaiteo Superiority: Normalo NonInferiority: Student to NonInferiority: Normalo Equivalency: Bivariate Student t• Ratio of Independent Meanso Superiority: Student to Superiority: Normalo NonInferiority: Student to NonInferiority: Normalo Equivalency: Bivariate Student t• Difference of Independent Means• (2x2 Crossover design)• Superiority: Student t• NonInferiority: Student t• Equivalency: Student t• Ratio of Independent Means (2x2 Crossover design)o Superiority: Student to NonInferiority: Student to Equivalency: Student t• Wilcoxon Mann Whitney Rank-Sum Test• Wilcoxon Mann Whitney Rank-Sum Test (2x2 Crossover Design)Many Means• One Way ANOVAo Equality of Many Meanso Single Contrast• One Way Repeated Measures ANOVA with Constant Correlationo Equality of Many Meanso Single Contrast• Two Way ANOVA:o Equality of Many MeansRegression• Linear Regression: Single Slope Coefficient• Linear Regression: Equality of Two Slope Coefficients• Repeated Measures Regression: Equality of Two Slope Coefficients• Multiple Linear Regression• Repeated Regression• Linear Mixed Effects Model: Difference of Means (Crossover Data)• Linear Mixed Effects Model: Ratio of Means (Crossover Data)Discrete• One Proportion• Single Proportiono Exact Testo Normal Approximation• McNemar's Test for Matched Pairs• Two Proportions• Difference of Independent Proportionso Superiorityo NonInferiorityo Equivalency• Ratio of Independent Proportionso Superiority• One Way ANOVAo NonInferiority Wald Testo NonInferiority Score Test• Odds Ratio of Independent Proportionso Superiorityo NonInferiority• Cocan Mantel Haenszel Test of a Common Odds Ratio in a Stratified 2x2 Table• Fisher's Exact Test• Many Proportionso Trend in R Ordered Proportionso Rx2 Chi-Square Test of Independence for Many Proportionso 1xC Chi-Square Test of Goodness of Fit for Many Proportionso 2xC Chi-Square Test of Independence for Many Proportionso RxC Chi-Square Test of Independence for Many Proportionso Wilcoxon Mann Whitney Rank-Sum Test for Ordered Categorical Data• Regressiono Logistic Regression : Single Slope Coefficient for a Normal Covariateo Logistic Regression : Single Slope Coefficient for a Normal Covariate Adjusted for Otherso Logistic Regressiono Probit Regressiono CLogLog RegressionAgreement• Cohen's Kappa Test of Two Proportions• Cohen's Kappa Test of Many ProportionsTime-To-Events• Two Survival Curves• Logrank Test of Equality of Survival Curveso Superiorityo NonInferiorityMultiple pairwise comparisons with a control:• Continuous endpoint: Difference of means• Binary endpoint: Difference of proportions• Parametric multiple comparison procedures:o Dunnett's single stepo Dunnett's step down• P-value based multiple comparison procedureso Bonferronio Sidako Weighted Bonferronio Holm's step downo Hochberg's step upo Hommel's step upo Fixed sequenceo Fallback• Generate means tough a DR curve• Global, disjunctive, conjunctive, and individual power• Family Wise Error Rate (FWER)• Comparative plots of power, FWER, box plots, estimated means3. Exact Statistical Methods. A statistical software package developed for use with exact statistical methods rather than asymptotic approximations is required. The functionality of the software should be available both as a standalone statistical package and in a SAS callable form. The following is list of methods and analyses which must be available:• One-sample Goodness-of-Fito Chi-Square o Kolmogorov o Lilliefors o Runs • Paired Sampleso Sign o Wilcoxon Signed-Rank o Hodges Lehman Estimates o Permutation o McNemar - Conditional o McNemar - Unconditional o Marginal Homogeneity • Two Independent Sampleso Wilcoxon-Mann-Whitney o Hodges Lehman estimates o Normal Scores o Savage Scores o Siegel-Tukey o Ansari-Bradley o Klotz o Mood o Conover o Permutation o Logrank o Wilcoxon-Gehan o Kolmogorov-Smirnov o Wald-Wolfowitz Runs • K Related Sampleso Friedman o Kendall's W o Cocan's Q o Quade o Page • K Independent Sampleso Median o Kruskall-Wallis o Normal Scores o Savage o ANOVA with General Scores o Jonckheere-Terpstra o Linear by Linear o Logrank o Wilcoxon-Gehan o Tarone and Ware Trend • One-Sample Rates and Proportionso Binomial o Blyth-Still- Cassella CI for binomial proportionso Blaker CI for binomial proportionso Multinomial o Poisson CI • Poisson Rateso Homogeneity of Relative Riskso CI on Common Relative o Trend in C Ordered Poisson Rates • Two Independent Binomialso Fisher's Exact o Pearson's Chi-square o Likelihood Ratio o CI on Odds Ratio o Barnard's Test for Superiorityo Tests of Non-inferiority o Tests of Equivalenceo CI on Difference of Proportions o CI on Ratio of Proportions • Two Related Binomialso McNemar's test o CI on odds Ratio o Test for Superiority o Tests of Non-inferiority o Tests of Equivalence o CI on Difference of Proportions • Stratified 2x2 Tableso Homogeneity of Odds Ratios o CI on Common Odds Ratios • C Ordered Binomials (with or without strata)o Cocan-Armitage Trendo Permutation with General Scores o Trend Test for Clustered Data o Test for Interaction Across Strata • Two Ordered Multinomials (with or without strata)o Wilcoxon-Mann-Whitney o Savage Scores o Normal Scores o Permutation with General Scores o Test for Interaction Across Strata • Unordered RxC Tableo Pearson's ChiSquare o Likelihood Ratio o Fisher-Freeman-Halton • Single Ordered RxC Tableo Kruskal-Wallis o Normal Scores o Savage o ANOVA with Arbitrary Scores • Doubly Ordered RxC Table• Jonckheere-Terpstra • Linear by Linear Association • Stratified RxC Tableso Unordered RxC Table o Single Ordered RxC Tableo Doubly Ordered RxC Table • Correlated Categorical Datao C Binomial Populationo Test of Correlationo Test for Trendo Multiple Binary Outcomeso Test for Trend • Measures of Association (nominal)o Contingency Coefficients o Goodman-Kruskal-Tau o Uncertainty Coefficient • Measures of Association (ordinal)o Pearson's Correlation o Spearman's Correlationo Kendall's Concordance o Kendall's Tau and Somers' D o Gamma Coefficient • Measures of Agreemento Cohen's Kappa o Weighted Kappa • Power & Sample Sizeo One Binomialo Paired Binomials:Difference o Two Binomials:Differenceo Superiority (conditional) o Non-inferiority o Equivalenceo K ordered Binomials • Two Ordered Multinomials (power) o Superiority (unconditional)o Two Binomials: Ratioo Non-inferiority Equivalenceo Superiority (unconditional)o Two Ordered Multinomialso Wilcoxon Ranksumo Savage Scoreso Normal Scoreso Permutation with General Score4. A software package/component for exact logistic regression, which includes the following analyses and capabilities:• Perform regression analysis for continuous, binary, polytonomous and count data. • Apply advanced regression techniques to data sets with missing values. • Accurately fit general linear models in cases of missing categorical covariates (models include Logit, Probit, CLoglog, Poisson and Normal.) • Handle large data sets using exact methods, Monte Carlo sampling and Markov Chain Monte Carlo sampling. • Methods to include: Firth's PMLE procedure; the calculation of mid-p corrected confidence intervals for a variety of models; best subset selection in binary logistic regression; the force inclusion of variable to the best subsets; and profile likelihood confidence intervals for parameters of binary logistic regression. • ROC curves for logistic models5. Specialized adaptive dose-finding trial design software package/component for Phase I/II designs, including validated statistical and simulation methods for the following:• Frequentist methodso Anovao Up and Downo t-test based Up and Downo Maximizing• Bayesian methodso Normal Dynamic Linear Model (NDLM)o Four Parameter Logistic (4PL)o Emax• R code support• Simulation tools• Visualization tools• Two-stage Bayesian and Frequentist designs for adaptively dropping doses • Early stopping for futility by conditional power in frequentist designs• Patient level simulation data • Plots o Dose response curves o Distribution of subjects o Subject response o Mean squared error o Estimated means/proportions o Target doses MCMC diagnostics o Credible intervalsComponents 1-5 may be separate software programs but should be interoperable and share a common user interface wherever possible.We do not have particular preferences for vendor selection. However, since the manufacturer sells at the same price to vendors and customers, we would expect a single bid result from putting the purchase out for bids.6.VENDOR RESPONSIBILITIESVendor will provide software, license files, upgrades, as well as annual maintenance and support. Delivery method can be either via CD/DVD or download.7.PERIOD OF PERFORMANCESeptember 7, 2016 - September 6, 2017 (base)September 7, 2017- September 6, 2018 (option year)8.0 Evaluation CriteriaThe Government will select the bidder whose offer best conforms to the solicitation and is most advantageous and provides the best value to the Government. The ability to provide the requested chemical simulation software and support is the most important factor. If proposals are equal in their technical merit, lowest price will become the determining factor.9.Instruction to quoters:The quoters shall submit a technical volume and a pricing volume. Each volume should include the following;Volume One Technical• Shall not exceed 10 pages in length• Lowest priced technically acceptable. Vendor can affirmatively demonstrate they can meet the requirements as stated in sections A-EVolume two pricing: • Unlimited number of pages• Separate pricing for base year with all required software• Separate pricing for option year one with all required softwareThe quoters shall include a completed copy of the provision FAR52212-3 Oferror Representation and Certifications-Commercial Items. ( October 2010) which can accessed@https://orca.bpn.govThe quoters shall comply with the provision at FAR 52.212-1(b) and submit and electronic copy to Karen Conroy (karen.conroy@fda.hhs.gov)Quoters responding to this RFP must be registered with SAM ( System Acquisition management)at http:www.sam.govQuoters must address the following FAR clauses FAR 52.212-4 Contract Terms and Condition Commercial Items, FAR 212-5 Contract Terms and Conditions Required to Implement Statutes or Executive Orders- Commercial Items. Far provision and clauses may be found at http://ww.gov/far/index.html10.Provisions and clausesSection 508 Accessibility Standards:• 1194.21 Software Publications and operating systems• 1194.31 Functional performance criteria• 1194.41 Information, Documentation, and Support.The following HHSAR clauses are incorporated by reference:• 352.201-Definitions ( Jan. 2006)• 352.270-12 Tobacco Free Facilities• 352-270-19 Electronic Information & Technology AccessibilityThese clauses may be downloaded fromhttp://www.hhs.gov/oamp/policies/hssar.doc52.211-6 Brand Name or Equal (Aug 1999) (a) If an item in this solicitation is identified as "brand name or equal," the purchase description reflects the characteristics and level of quality that will satisfy the Government's needs. The salient physical, functional, or performance characteristics that "equal" products must meet are specified in the solicitation. (b) To be considered for award, offers of "equal" products, including "equal" products of the brand name manufacturer, must- (1) Meet the salient physical, functional, or performance characteristic specified in this solicitation; (2) Clearly identify the item by- (i) Brand name, if any; and (ii) Make or model number; (3) Include descriptive literature such as illustrations, drawings, or a clear reference to previously furnished descriptive data or information available to the Contracting Officer; and (4) Clearly describe any modifications the quoters plans to make in a product to make it conform to the solicitation requirements. Mark any descriptive material to clearly show the modifications. (c) The Contracting Officer will evaluate "equal" products on the basis of information furnished by the quoter or identified in the offer and reasonably available to the Contracting Officer. The Contracting Officer is not responsible for locating or obtaining any information not identified in the offer. (d) Unless the quoter clearly indicates in its offer that the product being offered is an "equal" product, the quoter shall provide the brand name product referenced in the solicitation. (End of provision)FAR 52.217-9 Option to Extend the Terms of the Contract Submission must be received via email August 22, 2016 by 1200pm eastern standard time. Questions no later than August 17, 2016 . This is a firm fixed priced order.11.Place of Performance: FDA/Center for Drug Evaluation & Research10903 New Hampshire Ave. Silver Spring MD 2090312. Primary Point of Contact:Karen ConroyContract SpecialistKaren.conroy@fda.hhs.govPhone: 781-578-7452 fDA-SOL-16-1163455 Department of Health and Human Services Food and Drug Administration
8/12/16, 11:19 AM